Uploaded by 蔡菜

希腊罗马和近东

advertisement
Rome, the Greek World,
and the East
Studies in the History of Greece and Rome
Robin Osborne, P. J. Rhodes, and Richard J. A. Talbert, editors
Rome, the Greek World,
and the East
 
The Greek World, the Jews,
and the East
Fergus Millar
Edited by Hannah M. Cotton and Guy M. Rogers
The University of North Carolina Press
Chapel Hill
©  The University of North Carolina Press
All rights reserved
Set in Bembo
by Tseng Information Systems, Inc.
Manufactured in the United States of America
The paper in this book meets the guidelines
for permanence and durability of the Committee on
Production Guidelines for Book Longevity
of the Council on Library Resources.
Library of Congress Cataloging-in-Publication Data
Millar, Fergus.
Rome, the Greek world, and the East / Fergus Millar;
edited by Hannah M. Cotton and Guy M. Rogers.
p. cm.—(Studies in the history of Greece and Rome)
Includes bibliographical references and index.
Contents: v. . The Greek world, the Jews, and the East
-: ---- (cloth : alk. paper)
-: --- (cloth : alk. paper)
-: ---- (pbk. : alk. paper)
-: --- (pbk. : alk. paper)
. Greece—Civilization. . Rome—Civilization.
I. Title. II. Series.
 . 
—dc

cloth          
paper     
    
Contents
Preface
vii
Introduction to Volume , by Hannah M. Cotton
Abbreviations
xi
xix
Part I. The Hellenistic World and Rome
. The Problem of Hellenistic Syria

. The Phoenician Cities: A Case-Study of Hellenisation

. Hellenistic History in a Near Eastern Perspective:
The Book of Daniel

. The Background to the Maccabean Revolution: Reflections on
Martin Hengel’s ‘‘Judaism and Hellenism’’

. Polybius between Greece and Rome

. The Greek City in the Roman Period

Part II. Rome and the East
. Reflections on the Trials of Jesus

vi
Contents
. The Roman Coloniae of the Near East: A Study of Cultural Relations

. Latin in the Epigraphy of the Roman Near East

. Paul of Samosata, Zenobia, and Aurelian: The Church, Local Culture,
and Political Allegiance in Third-Century Syria

. Caravan Cities: The Roman Near East and Long-Distance
Trade by Land

. Looking East from the Classical World: Colonialism, Culture,
and Trade from Alexander the Great to Shapur I

Part III. Jews and Others
. Porphyry: Ethnicity, Language, and Alien Wisdom

. Hagar, Ishmael, Josephus, and the Origins of Islam

. Ethnic Identity in the Roman Near East, .. –:
Language, Religion, and Culture

. Dura-Europos under Parthian Rule

. The Jews of the Graeco-Roman Diaspora between Paganism
and Christianity, .. –

. Christian Emperors, Christian Church, and the Jews of the Diaspora
in the Greek East, .. –

Author’s Epilogue by Fergus Millar: Re-drawing the Map?

Index

Preface
Fergus Millar, Camden Professor of Ancient History in the University of
Oxford emeritus, is one of the most influential ancient historians of the
twentieth century. Since the publication of A Study of Cassius Dio by Oxford
University Press in , Millar has published ten books, including two
monumental studies, The Emperor in the Roman World (Duckworth, ) and
The Roman Near East,  ..–..  (Harvard, ). These books have transformed the study of ancient history.
In his study of the role of the emperor in the Roman World Millar argued
that the reign of Augustus inaugurated almost three centuries of relatively
passive and inert government, in which the central power pursued few policies and was largely content to respond to pressures and demands from below.
After nearly thirty years of scholarly reaction, The Emperor in the Roman World
is now the dominant scholarly model of how the Roman Empire worked in
practice.
Reviewers immediately hailed Millar’s magisterial study of the Roman
Near East as a ‘‘grand book on a grand topic’’ (TLS,  April ). In this
grand book, displaying an unrivaled mastery of ancient literary, epigraphic,
papyrological, and archaeological sources in Greek, Latin, Hebrew, Aramaic,
and other Semitic languages, Millar made the indigenous peoples of the Roman Near East, especially the Jews, central to our understanding of how and
why the three great religions of the book, Rabbinic Judaism, Christianity,
and Islam, evolved in a cultural context that was neither ‘‘eastern’’ nor ‘‘western.’’ There can be no doubt that The Roman Near East,  ..–..  will
be the standard work on the subject for a long time to come.
Over the past few years, Millar has published two books, The Crowd in
Rome in the Late Republic (Michigan, ) and The Roman Republic in Political
Thought (New England, ) on the politics of the Roman Republic and how
those politics have been understood or misunderstood by political thinkers
from the ancient world to the present. These books have challenged widely
vii
viii
Preface
held notions about the supposed oligarchic political character of the Roman
Republic. Most recently, based upon the Sather Lectures of  Millar now
has produced his first work centrally focused upon late antiquity, A Greek
Roman Empire: Power, Belief and Reason under Theodosius II, –.
In the future Millar intends to return to the Roman Near East, for a study
to be entitled Society and Religion in the Roman Near East from Constantine to
Mahomet. In this study Millar will bring the story of Greco-Roman culture
in the Near East from the early fourth century up to the Islamic invasions of
the seventh century .. Apparently, the enormous task of writing a social
history of the Near East from Alexander to Mahomet will not be left completely to someone else after all, as Millar prophesied in  (The Roman Near
East, pp. xii–xiii).
During the same period when he has produced these ground-breaking
books, Millar also has published more than seventy essays on aspects of
Greco-Roman history, from the Hellenistic period until the middle of the
fifth century .. These essays have laid the foundations for or supplemented
the ideas and arguments presented in Millar’s very well known books. Some
of these essays, such as ‘‘The Emperor, the Senate and the Provinces’’ ( Journal of Roman Studies  []: –), or ‘‘Emperors, Frontiers and Foreign Relations,  ..–.. ’’ (Britannia  []: –), have appeared in
hitherto accessible journals and are widely regarded as classics of scholarship. But other outstanding essays, such as Millar’s study ‘‘Polybius between
Greece and Rome’’ (published in Greek Connections: Essays on Culture and Diplomacy [], –), have been more difficult to locate, even for professional
historians doing research in the field.
Therefore, the primary goal of our collection, Rome, the Greek World, and
the East, is to bring together into three volumes the most significant of Millar’s essays published since  for the widest audience possible. The collection includes many articles that clearly will be of great intellectual interest
and pedagogical use to scholars doing research and teaching in the different
fields of the volume headings: Volume , The Roman Republic and the Augustan
Revolution; Volume , Government, Society, and Culture in the Roman Empire;
and Volume , The Greek World, the Jews, and the East.
At the same time, we have conceived and organized the three volumes of
Rome, the Greek World, and the East especially in order to make Millar’s most
significant articles readily available to a new generation of students, who increasingly may not have access to the specialty journals or edited volumes in
which many of Millar’s more recent articles have appeared.
The principle of arrangement of the essays in each of the three volumes
is broadly chronological by subject matter treated within the ancient world.
Preface
ix
We believe that this chronological arrangement of essays (rather than by publication date of the essays) gives intellectual coherence to each volume on its
own and to the collection as a whole. Overall, as Millar himself has defined
it, the subject of this collection is ‘‘the communal culture and civil government of the Graeco-Roman world, essentially from the Hellenistic period
to the fifth century ..’’ (‘‘Author’s Prologue,’’ volume , p. ).
Publication of a three-volume collection of essays, drawn from a wide
variety of journals and edited volumes, over more than four decades of scholarly production, presents editors with some major stylistic challenges. Our
collection contains fifty-four essays (in addition to a prologue and an epilogue written by Millar especially for the present collection). Most of these
essays originally were published in learned journals or books, each of which
had its own house style. Some learned journals also have changed their house
styles over the time when Millar has published in them. For these reasons we
have not attempted to bring all of the citations in the texts or notes of the
articles in the collection into perfect stylistic conformity. Conformity for the
sake of conformity makes no sense; moreover, to achieve such conformity
would delay publication of the collection for years.
Rather, the stylistic goal of our collection has been to inform readers
clearly and consistently where they can find the sources cited by Millar in his
essays. To help achieve that goal we have included a list of frequently cited
works (with abbreviations for those works) at the beginning of each volume.
Thus, in the text or notes of the essays, readers will find abbreviations for
frequently cited journals or books, which are fully cited in our lists at the
beginning of each volume. For example, references in the notes to the abbreviation JRS refer to the Journal of Roman Studies. For the abbreviations themselves we have relied upon the standard list provided in L’Année Philologique.
In certain cases, where there have been individual citations in the original
texts or notes to more obscure collections of inscriptions or papyri, we have
expanded the citations themselves in situ, rather than endlessly expanding
our list of frequently cited works.
In accordance with Fergus Millar’s wishes, for the sake of readers who do
not know Latin or Greek, we have provided English translations of most of
the extended Greek and Latin passages and some of the technical terms cited
by Millar in the text and notes of the original essays. In doing so, we have followed the practice Fergus Millar himself adopted in The Emperor in the Roman
World in  and in The Roman Near East in . We believe that providing
these translations will help to make Millar’s essays more widely accessible,
which is the essential goal of the collection. Readers who wish to consult the
original Greek, Latin, and Semitic passages or technical terms that we have
x
Preface
translated in the collection can look up those passages or technical terms in
the original, published versions of the essays.
The editors would like to thank the many friends and colleagues who have
helped us in the process of collecting these essays and preparing them for
publication. We are indebted first of all to Lewis Bateman, formerly senior
editor at the University of North Carolina Press, who suggested the basic arrangement of the essays into three volumes. We are also grateful to David
Perry, editor-in-chief; Charles Grench, assistant director and senior editor;
Pamela Upton, assistant managing editor; Amanda McMillan, assistant editor; and last but certainly not least, Brian MacDonald, our copyeditor, at the
University of North Carolina Press, for their flexibility, advice, and support
of the project.
Asaph Ben Tov, Masha Chormy, Tamar Herzig, Amir Marmor, Andrea
Rotstein, Naomi Schneider, and Ori Shapir of the Hebrew University in
Israel and Michal Molcho in Oxford retyped, converted, scanned, and corrected old and new articles into new format, thereby providing the editors
with excellent manuscripts to work on. Dr. Nancy Thompson of the Metropolitan Museum of Art in New York provided invaluable editorial assistance
at all stages. Our thanks also to Mark Rogers for his help with the maps. We
owe a very unique debt to Priscilla Lange whose generosity, kindheartedness, and hospitality in Oxford has made all the difference. We also would
like to express our gratitude to the Fellows of Brasenose College Oxford and
All Souls College Oxford for their hospitality while we were working on this
project.
Above all, however, the editors would like to thank Fergus Millar, for his
scholarship, for his generosity, and for his friendship over three decades.
Hannah M. Cotton
The Hebrew University
Jerusalem
Guy M. Rogers
Wellesley College
Wellesley
Introduction
The Greek World, the Jews, and the East is the third and last volume in the series
of Fergus Millar’s collected essays, Rome, the Greek World, and the East. It stands
to The Roman Near East,  ..–..  1 as the second volume in the series
stands to The Emperor in the Roman World,  ..–.. ,2 and, to a lesser
extent, as the first volume in the series stands to The Crowd in Rome in the Late
Republic.3
The two themes of the first volume, The Roman Republic and the Augustan
Revolution (), are the democratic nature of the Roman Republic and the
transition from Republic to Principate, and its focal point, the cradle of the
empire—the city of Rome and Italy. The second volume, Government, Society,
and Culture in the Roman Empire (), is about imperial Rome; it deals with
the empire as a system of government; only briefly and intermittently does
it touch on local cultures in the provinces. The present volume is a study-indepth of the impact of empire on the eastern part of the Empire. The ‘‘East’’
in the title is the Roman Near East, which eventually extended much further
east than it had in the early Empire. With the incorporation of great tracts
of the East, Rome ceased, also in the East, to be a Mediterranean empire.
Millar’s Roman Near East is that part of the Roman world populated
by Semitic-speaking and Semitic-writing peoples;4 their encounter with
. Based on the Carl Newell Jackson Lectures (Harvard, ), and published by Harvard
University Press, , and in paperback,  (henceforth RNE ).
. Duckworth and Cornell University Press, , and a second edition with an afterword,  (henceforth ERW ).
. Based on the Jerome Lectures (Ann Arbor, Michigan, autumn , and the American
Academy in Rome, ) and published by Michigan University Press, .
. Egypt is absent from this volume as it is essentially absent from RNE; cf. Millar’s epilogue in this volume (text to n. ). Nonetheless, observe the caveats in the prologue to RNE
against facile identification of language and ethnicity, and the assumption of a common
Semitic culture.
xi
xii
Introduction
Greco-Roman civilization is the leitmotif running through all of the articles
in this volume. The study of ethnicity and language accounts for an apparently almost obsessive preoccupation with primary evidence, inscriptions
and papyri.
I can offer no better description of the conception behind RNE and therefore also behind the articles collected here than that given by Fergus Millar
himself in chapter  (‘‘Porphyry: Ethnicity, Language, and Alien Wisdom,’’
) of the present volume, where he debunks the notion of ‘‘something
distinctively ‘oriental’ in his thought,’’ ascribed to the Neoplatonist philosopher solely because he was born in Tyre in s:
The Roman Near East,  ..–..  () . . . was intended precisely to draw a ‘‘map’’ of the Roman provinces of Syria, Mesopotamia, Judaea/Syria Palaestina, and Arabia, first in terms of language,
as attested by documentary and literary evidence, and then of ‘‘ethnicity,’’ as claimed or imputed by contemporaries. Its approach was thus
severely empirical, and was designed, not—as is in any case impossible—to prove negatives (for instance, that a particular language was
not in use at a particular time or place), but to show the limits of what
we know now: that we have actual evidence for the use of a particular
language only in particular places and at particular times. Suppositions
about wider use are themselves hypotheses, which might of course receive support at any time from new evidence. But such suppositions,
until supported, must on no account serve as the bases for interpreting
the thought of Porphyry, or of other philosophers.
Many of the chapters in the present volume, albeit self-contained, are indeed
preliminary studies for RNE, as their first footnote attests; others, looking
backward to the Hellenistic period, precede the time span set by RNE; some,
written after the completion of that work, look forward to its sequel; finally,
the two last chapters go beyond RNE ’s thematic scheme to discuss the Jewish
diaspora in the West as well as in the East.
However, each of the chapters has a focal point, which in one way or another has been integrated into the vast canvas of RNE, or at least has been
foreshadowed in it. Their innovativeness, originality, and individuality are
thus in danger of being lost to the unwary and inexperienced reader of Fergus Millar’s work—which now spans almost half a century. It is one of the
aims of the present collection, as it was of the previous two, to recapture
the vigor and ground-breaking freshness of the originals for a young and
uninitiated audience and recreate it for others.
The six chapters of the first part, ‘‘The Hellenistic World and Rome,’’ be-
Introduction
xiii
long to the Hellenistic period, before RNE begins, yet it would be no exaggeration to say that the first chapter, ‘‘The Problem of Hellenistic Syria’’
(), has it all in a nutshell. Here is struck the first chord of dissent from and
challenge to widely accepted and much venerated traditions. Millar’s empirical approach needed no more than a few strokes to sweep away numerous
comfortable views on the time-honored concept of Hellenization as fusion
(Verschmeltzung) of Greek and local cultures. ‘‘The ‘Hellenistic’ Syria, with a
distinctive mixed culture’’ turns out to be a product of its Roman phase, ‘‘that
which evolved under the Roman Empire’’ (chapter , text to n.  below).
The advent of Rome brought with it considerable intensification of the process of Hellenization as attested in the written corpus as well as in material
culture. Until then our evidence for the Greek presence in Syria is ‘‘limited, variable, and erratic.’’ Even more disconcerting is the piercing question:
‘‘the Hellenisation of what?’’ What do we know of the Achaemenid period
in Syria? This is taken up again in RNE: ‘‘So far as our evidence goes, the
preceding Hellenistic period has left almost nothing which can count as the
expression of a regional or a local culture’’ (). As it happens the indigenous
local cultures too owe their very appearance in the historical record to their
adoption of the ‘‘epigraphic habit’’ of the Greco-Roman world, mostly in
Greek, but sometimes in their own languages, under Roman sway. A tempting solution to ‘‘the amnesia of the Semitic cultures’’ is, of course, the model
offered by Glen Bowersock, namely that ‘‘one of the functions of GraecoRoman culture in this region may precisely have been to offer a vehicle for
the expression of an indigenous local culture’’ in the area.5
The counterpoint to this is the subject of chapter : ‘‘The Phoenician
Cities: A Case-Study of Hellenisation’’ (), where true fusion of indigenous and Greek cultures in the old sense seems to have taken place. However,
the Hellenization of the Phoenician cities began long before Alexander and
has rarely been discussed in this context. A Jewish version of the fusion is
the book of Daniel (chapter : ‘‘Hellenistic History in a Near Eastern Perspective: The Book of Daniel,’’ ), written in the s .., which may
borrow from Hellenism ‘‘the notion of history as a succession of world empires,’’ but it does so only in order to reaffirm Jewish monotheism against the
Hellenistic persecutor Antiochus IV Epiphanes. Neither should the events
of – .. be seen as the culmination of a profound and long process
of evolution within Judaism, assimilating it to Hellenism (chapter : ‘‘The
Background to the Maccabean Revolution: Reflections on Martin Hengel’s
. Millar’s phrasing in RNE, , referring to G. R. Bowersock, Hellenism in Late Antiquity (Thomas Spencer Jerome Lectures, ), esp. chap. .
xiv
Introduction
‘Judaism and Hellenism,’ ’’ ). We should strongly resist the tendency of
modern historiography to search for deeper reasons and ‘‘look to sources
other than the reigning Seleucid king himself for the explanation of a persecution carried out by royal command and by royal agents’’ (chapter , text
to n. ; cf. RNE, ). Instead, we should see the desecration of the Temple
in Jerusalem for what it was: in no sense syncretistic but the arbitrary imposition of a pagan cult.
The reader of chapter , ‘‘Polybius between Greece and Rome’’ (),
where Millar denies the attribution of Roman sympathies to the historian,
cannot but recall Millar’s first book, A Study of Cassius Dio (). In the span
of time separating the two, a Greek historian not only ceased to regard Rome
as the invader, but could become Roman himself while ‘‘staying Greek’’—
a fusion of cultures that could rightly be described as Hellenism. In similar
vein, the Greek city of the imperial period (chapter , ‘‘The Greek City in
the Roman Period,’’ ) ‘‘would be more correctly described as ‘GraecoRoman’: that is, as a fusion or mélange of languages and constitutions, types
of public entertainment, architectural forms, and religious institutions’’ (text
to n. ), detached from its place of birth, Greece and Asia Minor, and transcending even the borders of the Empire—‘‘a symbol of the fact that in the
end Graecia capta did indeed imprison her captor.’’
The second part is rather heterogeneous and the least cohesive of the three,
but it is far from chaotic. All six chapters reveal the complex and subtle impact of the Roman presence in the East, never to be dismissed, never to be lost
to sight, but always working in intricate and unexpected ways. Each of them
raises problems to which solutions and interpretations have been offered in
the past, solutions and interpretations that are now shown to be less than
satisfactory and sometimes patently wrong.
Chapter , ‘‘Reflections on the Trials of Jesus’’ (), is dedicated to Geza
Vermes, coeditor with Fergus Millar of the new Schürer.6 Of all the chapters in this volume, this is the most likely to give the reader a taste of what
it was like to be Fergus Millar’s student, to take lessons from him in how to
read an ancient text and in how to choose the more veridical version of the
course of events. In dealing with the trial of Jesus, ‘‘we must not proceed by
amalgamating data from all four gospels’’ nor, surprisingly, by weighing arguments based on coherence and plausibility, but rather by choosing the one
that has the better grasp ‘‘on the realities of Palestine under Roman domination,’’ which in this case means the one that conforms ‘‘with the world as
. E. Schürer, G. Vermes, and F. Millar, The History of the Jewish People in the Age of Jesus
Christ,  ..–.. , vols. I–III (Edinburgh, –).
Introduction
xv
portrayed by Josephus.’’ The preference, which is ‘‘no more than an hypothesis,’’ goes to John’s Gospel.
The next two chapters belong together for two reasons. First, a formal
reason: it was our editorial decision, contrary to our practice throughout,
not to translate the texts written in the ancient languages in these two chapters. These are in essence linguistic papers based on documents, focusing on
borrowings from one language to another. Although integral to the central theme of the volume, these two chapters do not, and can not, make
sense unless the relevant material is retained in the original (or in the case of
Semitic languages in upper-case transliteration). To ‘‘transcribe’’ and translate
this material would blur the intellectual point.
Second, both chapters isolate and explore the Roman component in the
mixture of cultures characteristic of the Roman Near East and its impact.
This is seen in the foundation of colonies or the conferral of the title colonia
on native cities (chapter : ‘‘The Roman Coloniae of the Near East: A Study
of Cultural Relations,’’ —virtually a monograph), and in the limited use
of the Latin language (chapter : ‘‘Latin in the Epigraphy of the Roman Near
East,’’ ). The bilingual and trilingual documents from Palmyra discussed
in chapter  hammer home the point made by Millar already in his celebrated article on ‘‘Epigraphy,’’ 7 namely that the richness of the epigraphic
tradition comes fully into its own only when epigraphic texts in different
languages, the contemporaneous expressions of different but related cultures,
are studied together.8
A test case for the survival and potential vitality of local cultures underneath the dominant Greek surface is the subject of chapter , ‘‘Paul of Samosata, Zenobia, and Aurelian: The Church, Local Culture, and Political Allegiance in Third-Century Syria’’ (), which examines the hypothesis that
in the events of the s and the s in Roman Syria ‘‘either a man from
Samosata or a ruler of Palmyra could have seen themselves as in any sense
representatives of the ‘Orient’ as against the Graeco-Roman world.’’
The next two chapters should also be read together. They explore the true
‘‘Orient,’’ the one beyond the Euphrates and outside the classical world, in
. Originally published in  chapter  in F. Millar, Rome, the Greek World, and the
East I: The Roman Republic and the Augustan Revolution (), ; cf. section  of the epilogue
to this volume: ‘‘A New Approach to Ancient Languages?’’
. The innovative project now under way to create a comprehensive multilingual corpus of all inscriptions, both published and unpublished, from the territory of presentday Israel, from the fourth century ... to the seventh century .. (Corpus Inscriptionum
Iudaeae/Palaestinae CIIP ) owes Millar a great debt; see Cotton et alii, ‘‘Corpus Inscriptionum Iudaeae/Palaestinae,’’ ZPE  (), –.
xvi
Introduction
two ways. First, we follow the possibility of the existence of long-distance
overland trade with central Asia, with its necessary corollary, caravan cities
in the sense used by Millar’s favorite historian, M. I. Rostovtzeff, and with
consequences for our understanding of ‘‘the ancient economy’’ (chapter :
‘‘Caravan Cities: The Roman Near East and Long-Distance Trade by Land,’’
; cf. RNE, ). And, second, we are taken on a kaleidoscopic journey
through the six centuries following Alexander’s conquests, tracing the impact of Greco-Roman culture in the Asian land-mass (chapter : ‘‘Looking
East from the Classical World: Colonialism, Culture, and Trade from Alexander the Great to Shapur I,’’ ).
The third part, ‘‘Jews and Others,’’ revolves around ethnicity and selfdefinition of various groups. Chapter  (‘‘Porphyry: Ethnicity, Language,
and Alien Wisdom,’’ ) has already been mentioned, and we may pass on
to chapter , dedicated to the memory of Menahem Stern, author of the
invaluable Greek and Latin Authors on Jews and Judaism ( vols., Jerusalem, –
),9 whose life and work reaffirm Millar’s profound belief in the lasting
strength of the ‘‘national, or ethnic, historical and cultural traditions’’ of the
Jewish people. ‘‘The Arabes or Arabioi of antiquity have so far not had a Menahem Stern to collect and analyse all the classical references to them, which
begin in the fifth century with Aeschylus and Herodotus’’ (chapter , text to
n. ). Their legendary common genealogy, however, was not supplied within
Greco-Roman mythology but, ironically, within Jewish biblical history, by
none other than the Jewish Josephus who saw in them the descendants of
Ishmael, son of Abraham and Hagar, thereby making them legitimate heirs
to Jewish monotheism, with mighty consequences for the history of Islam
(‘‘Hagar, Ishmael, Josephus, and the Origins of Islam,’’ ; cf. RNE, ).
Chapter  (‘‘Ethnic Identity in the Roman Near East, .. –: Language, Religion, and Culture,’’ ) encapsulates succinctly the first part of
the planned sequel to RNE, taking the story of ethnicity, religion, and language down to the middle of the fifth century. The continuing predominance of Greek in this network of Greek cities and in their territories, despite the gradual emergence of Syriac within the Christian church, leaves the
complex and fascinating issues of ethnicity, communal identity, and religious
belief outside the lines of linguistic demarcation.
It is to be hoped that the challenge thrown down by Millar in chapter , ‘‘Dura-Europos under Parthian Rule’’ (), will be picked up and
the rich epigraphic, papyrological, and architectural evidence will be used to
write the history of Hellenistic, Parthian, and Roman Dura. At the very least,
. Volume  was reviewed by Millar in SCI  (): –.
Introduction
xvii
the legal documents, their formulae, diplomatics, and language, should be
studied (chapter , text to notes –) and compared with their counterparts from other sites in the Roman Near East—the documents from the
Judean Desert and the archives from the Euphrates, from Nessana, and now
from Petra.10
The last two chapters (chapter : ‘‘The Jews of the Graeco-Roman Diaspora between Paganism and Christianity, .. –,’’ , and chapter :
‘‘Christian Emperors, Christian Church, and the Jews of the Diaspora in the
Greek East, .. –,’’ ) are about the Jewish diaspora. Not that the
Jews are ever out of sight elsewhere in this volume or in RNE. An example
is the typical statement in the introductory pages to chapter :
For what distinguished the Jews from all the other groups which came
within the orbit of Graeco-Roman civilisation was precisely the fact
that they already possessed a national religious history, the Bible (even
if not yet quite complete in its canonical form), written in a literary
language entirely independent of Greek or Latin. Existing as a corpus
of literary works, the Bible could be and was translated into Greek in
the Hellenistic period.
It is legitimate to ask if the same holds true for Judaism and the Jews in the
diaspora as for those of Palestine. At any rate in the period discussed in these
last two chapters, in Palestine as elsewhere the Jews, now no longer a majority, were caught between Christianity and a declining paganism, and the
issues cannot have been fundamentally different from those outside. What is
beyond dispute after reading these two chapters is ‘‘that the history of Judaism in late antiquity cannot be confined to Palestine, for we can see that there
was a very significant Jewish life in the diaspora’’ (chapter , last paragraph).
With this the third and last volume in the series comes to an end, but not
completely; we can trust Fergus Millar not to sit idle when ‘‘there is work to
be done.’’ The publication of articles continues.
We may fittingly end with the traditional Jewish exhortation on finishing
the reading of each book of the Pentateuch: ḤZQ ḤZQ WNTḤZQ, which
can be loosely translated as ‘‘be strong and of good courage.’’
Hannah M. Cotton
Jerusalem
 July 
. Cf. H. M. Cotton, W. E. H. Cockle, and F. G. B. Millar, ‘‘The Papyrology of the
Roman Near East: A Survey,’’ JRS  (): –.
This page intentionally left blank
Abbreviations
AAAS
Annales Archéologiques Arabes Syriennes
AAES, Ann. Arch. Ar. Syr.
American Archaeological Expedition to Syria
Abbott and Johnson, Municipal Administration
F. F. Abbott and A. C. Johnson, Municipal Administration in the Roman
Empire ()
Abh. d. Kön. Ges. der Wiss. zu Göttingen
Abhandlungen der Königlichen Gesellschaft der Wissenschaften zu Göttingen
ABSA
Annals of the British School at Athens
AC
L’Antiquité Classique
ADAJ, Ann. Dept. Ant. Jordan
Annual of the Department of Antiquities of Jordan
AE, Ann. Épig.
L’Année Épigraphique
AJA, Am. Journ. Arch.
American Journal of Archaeology
AJAH
American Journal of Ancient History
AJBA
Australian Journal of Biblical Archaeology
AJPh
American Journal of Philology
Am. Num. Soc. Mus. Notes
American Numismatic Society, Museum Notes
Anat. Stud.
Anatolian Studies
xix
xx
Abbreviations
Ann. Sc. N. Sup. Pisa
Annali della Scuola Normale Superiore di Pisa
ANRW
Aufstieg und Niedergang der römischen Welt
BAR Int. Ser.
British Archaeological Reports, International Series
BASOR
Bulletin of the American Schools of Oriental Research
BCH
Bulletin de Correspondance Hellénique
BE, Bull. Épig.
Bulletin Épigraphique, published in Revue des Études Grecques
BEFAR, Bib. Ec. Fr. Ath. Rom.
Bibliothèque des Écoles Françaises d’Athènes et de Rome
BES
Bulletin d’Épigraphie Sémitique, published in Syria
BGU
Aegyptische Urkunden aus den Königlichen (later Staatlichen) Museen zu
Berlin, Griechische Urkunden
BIAB
Bulletin de l’Institut Archéologique Bulgare
Bibl. Arch. Rev.
Biblical Archaeology Review
BICS
Bulletin of the Institute of Classical Studies of the University of London
BJ, Bonn. Jahrb.
Bonner Jahrbücher des Rheinischen Landesmuseums in Bonn und des Vereins
von Altertumsfreunden im Rheinlande
BMC Arabia, BMC Mesopotamia
G. F. Hill, Catalogue of the Greek Coins of Arabia, Mesopotamia and Persia in
the British Museum
BMC Palestine
G. F. Hill, Catalogue of the Greek Coins of Palestine in the British Museum
BMC Phoenicia
G. F. Hill, Catalogue of the Greek Coins of Phoenicia in the British Museum
BMC Syria
W. W. Wroth, Catalogue of the Greek Coins of Galatia, Cappadocia and Syria
in the British Museum
Bull. Mus. Beyrouth
Bulletin du Musée de Beyrouth
Abbreviations
xxi
Bull. Sch. Or. and Afr. Stud.
Bulletin of the School of Oriental and African Studies
Bull. Soc. Ant. Alex.
Bulletin de Société des Antiquaires d’Alexandria
Bull. Soc. Nat. Ant. France Bulletin
de la Société Nationale des Antiquaires de la France
CAH
Cambridge Ancient History
CCL, Corpus Christ. Lat., Corp. Christ. Lat.
Corpus Christianorum, Series Latina
Chron. d’Ég., CE
Chronique d’Égypte
CIG
Corpus Inscriptionum Graecarum
CIL
Corpus Inscriptionum Latinarum
CIS
Corpus Inscriptionum Semiticarum
Class. Phil., CPh
Classical Philology
Corp. Ins. Jud., CIJ
J.-B. Frey, Corpus Inscriptionum Iudaicarum. Recueil des inscriptions juives qui
vont du III e siècle avant Jésus-Christ au VII e siècle de notre ère I–II (–)
Corpus. Gloss. Lat.
Corpus Glossariorum Latinorum
CPJ
Corpus Papyrorum Iudaicarum I–III
CQ
Classical Quarterly
CR
Classical Review
CRAI
Comptes-rendus de l’Académie des Inscriptions
CSCO
Corpus Scriptorum Christianorum Latinorum
CSEL
Corpus Scriptorum Ecclesiasticorum Latinorum
Dam. Mitt.
Mitteilungen des Deutschen Archäologischen Instituts, Damaszener Abteilung
xxii
Abbreviations
Devijver, Pros. milit. Equestr.
H. Devijver, Prosopographia Militiarum Equestrium quae fuerunt ab Augusto
ad Gallienum I–V (–)
Dial. di Arch.
Dialoghi di Archeologia
Diz. Epig.
Dizionario Epigrafico
DJD
Discoveries in the Judaean Desert I–XXIX (Oxford, –)
EHR, Econ. Hist. Rev.
Economic History Review
Eph. Theol. Lovanienses
Ephemerides Theologiae Lovanienses
Epigr. Anat.
Epigraphica Anatolica
FGrHist., FGrH
F. Jacoby, Die Fragmente der griechischen Historiker
FHG
C. Müller, Th. Müller, et al., Fragmenta Historicorum Graecorum I–V
(–)
GGM
Geographi Graeci Minores, ed. C. Müller (–)
GRBS, Gr. Rom. and Byz. St.
Greek, Roman and Byzantine Studies
HSCPh, HSCP, Harv. Stud. Class. Phil.
Harvard Studies in Classical Philology
HThR, Harv. Th. Rev.
Harvard Theological Review
ID, Ins. Délos
Inscriptions de Délos (–, )
IEJ
Israel Exploration Journal
IG
Inscriptiones Graecae
IGBulg.
Inscriptiones Graecae in Bulgaria repertae
IGLS, Ins. Gr. et Lat. de la Syrie
Inscriptions Grecques et Latines de la Syrie
IGR
Inscriptiones Graecae ad Res Romanas Pertinentes
Abbreviations
xxiii
IJudO
Inscriptiones Judaicae Orientis I: Eastern Europe, ed. D. Noy, A. Panyatov,
and A. Bloedhorn; II: Kleinasien, ed. W. Ameling; III: Syria and Cyprus,
ed. D. Noy and H. Bloedhorn ()
I. K. Alexandria Troas
Inschriften griechischer Städte aus Kleinasien: Alexandria Troas
I. K. Central Pisidia
Inschriften griechischer Städte aus Kleinasien: Central Pisidia
I. K. Lampsakos
Inschriften griechischer Städte aus Kleinasien: Lampsakos
ILS
Inscriptiones Latinae Selectae I–III
Ins.
Didyma II, Die Inschriften
Inventaire, Inv.
Inventaire des Inscriptions de Palmyre I–, ed. J. Cantineau (–)
IRT
Inscriptions of Roman Tripolitana
Isr. Num. Journ.
Israel Numismatic Journal
Ist. Mitt.
Mitteilungen des Deutschen Archäologischen Instituts, Istanbuler Abteilung
Jahrb. d. Öst. Byz. Ges.
Jahrbuch der Österreichischen Byzantinischen Gesellschaft
Jahrb. f. Ant. u. Chr.
Jahrbuch für Antike und Christentum
JBL, Journ. Bib. Lit.
Journal of Biblical Literature
JEA
Journal of Egyptian Archaeology
JHS, Journ. Hell. Stud.
Journal of Hellenic Studies
JJS, Journ. Jew. Stud.
Journal of Jewish Studies
JNES
Journal of Near Eastern Studies
Journ. Med. Stud.
Journal of Mediterranean Studies
JQR
Jewish Quarterly Review
xxiv
Abbreviations
JRA
Journal of Roman Archaeology
JRS, Journ. Rom. Stud.
Journal of Roman Studies
JS
Journal des Savants
JSJ
Journal for the Study of Judaism
JSS
Journal of Semitic Studies
JThS, JThSt, Journ. Theol. Stud., J. Theol. St., J. Th. St.
Journal of Theological Studies
KAI
Kanaanaïsche und Aramaïsche Inschriften 3, ed. H. Donner and W. Rölling
(–)
MAMA
Monumenta Asiae Minoris Antiqua
Med. Hist. Rev.
Mediterranean Historical Review
Mél. Univ. St. Joseph, MUSJ
Mélanges de l’Université Saint-Joseph
Mém. Miss. Arch. Iran
Mémoires de la Mission Archéologique Française en Iran
Mém. Soc. Nat. Ant. Fr.
Mémoires de la Société Nationale des Antiquaires de France
Migne, PG
See PG
MUSJ
See Mél. Univ. St. Joseph
NC, Numis. Chron.
Numismatic Chronicle
Num. Notes and Monog.
Numismatic Notes and Monographs
OA
Oriens Antiquus
OGIS
Orientis Graeci Inscriptiones Selectae I–II
P.Cair.Zen., PCZ
Zenon Papyri, Catalogue Général des Antiquités Égyptiennes du Musée du
Caire I–V
Abbreviations
xxv
P. Dura
The Excavations at Dura-Europos conducted by Yale University and the French
Academy of Inscriptions and Letters, Final Report V, Part I, The Parchments
and Papyri, ed. C. B. Welles, R. O. Fink, and J. F. Gilliam ()
P. Lond.
Greek Papyri in the British Museum
P. Oxy.
The Oxyrhynchus Papyri
PAES
Princeton Archaeological Expedition to Syria
PAT
Palmyrene Aramaic Texts, ed. D. R. Hillers and E. Cussini ()
PCPS, Proc. Camb. Phil. Soc., PCPhS, PCPS, Proc.
Proceedings of the Cambridge Philological Society
PEQ
Palestine Exploration Quarterly
PG
Patrologia Graeca
PL
Patrologia Latina
PO, Patr. Or.
Patrologia Orientalis
Proc. Brit. Acad.
Proceedings of the British Academy
P-W, RE
Pauly-Wissowa, Realencyclopädie der Klassischen Altertumswissenchaft
QDAP
Quarterly of the Department of Antiquities in Palestine
RAC
Reallexikon für Antike und Christentum
RB
Revue Biblique
RE
See P-W
Rec. Arch. Or.
Recueil d’Archéologie Orientale
REG
Revue des Études Grecques
Rev. Ét. Juives
Revue des Études Juives
xxvi
Abbreviations
Rev. hist. relig., Rev. Hist. Rel., RHR
Revue de l’Histoire des Religions
Rev. Num., RN
Revue Numismatique
Rev. Or. Chr.
Revue de l’Orient Chrétien
Rev. Phil.
Revue Philologique
Rev. Sci. Rel.
Revue de Sciences Religieuses
Riv. Arch. Cr.
Rivista di Archeologia Cristiana
RPC
The Roman Provincial Coinage I–II, ed. A. Burnett, M. Amandry, and P. P.
Ripollès (–)
Schürer, Vermes, and Millar, History
E. Schürer, G. Vermes, and F. Millar, The History of the Jewish People in the
Age of Jesus Christ,  ..–..  I–III (–)
SCI
Scripta Classica Israelica
SDB
Supplément au Dictionnaire de la Bible
SEG
Supplementum Epigraphicum Graecum
Sel. Pap.
Select Papyri. The Loeb Classical Library
SNG
Sylloge Nummorum Graecorum
Stern, Greek and Latin Authors
M. Stern, Greek and Latin Authors on Jews and Judaism I–III (–)
Syll.3
Sylloge Inscriptionum Graecarum I–IV 3
TAPhA
Transactions of the American Philological Association
TLS
Times Literary Supplement
YCS, Yale Class. Stud.
Yale Classical Studies
ZDMG
Zeitschrift der Deutschen Morgenländischen Gesellschaft
Abbreviations
xxvii
ZDPV
Zeitschrift des Deutschen Palästina-Vereins
ZPE, Zeitschr. f. Pap. u. Epig.
Zeitschrift für Papyrologie und Epigraphik
ZSS, ZRG
Zeitschrift der Savigny-Stiftung für Rechtsgeschichte, Romanistische Abteilung
This page intentionally left blank
Rome, the Greek World,
and the East
The Roman Empire, ca.  ..
This page intentionally left blank
PA RT I
The Hellenistic World and Rome
This page intentionally left blank
 
The Problem of Hellenistic Syria *
And it came to pass after the victory of Alexander the son of Philip, the
Macedonian, who came out from the land of Kittim and smote Darius,
king of the Persians and Medes . . . and started many wars and conquered many fortified places and slew the kings of the earth. . . . And his
sons ruled, each in his own place, and after his death they all assumed
diadems, and his sons (ruled) after him for many years and multiplied
evils in the land.
— Maccabees .–
The first book of Maccabees in its opening paragraph reflects an important
aspect of the impact of Hellenistic rule in Syria, the prevalence of conflict,
war, and instability. It does also, however, illustrate something quite different, the possibility of a communal historical consciousness and a national
culture which might provide a framework within which a community in
* First published in A. Kuhrt and S. M. Sherwin-White, eds., Hellenism in the East (London,
), –.
The work on which this survey of the problem of Hellenistic Syria is based was carried
out at the Institute for Advanced Study at Princeton, during an enjoyable and profitable
visit from January to April . Successive versions were presented at seminars held at the
institute and, in autumn , at the Institute of Classical Studies, London. The paper has
benefitted from assistance, advice, and criticism from the editors and from a number of
friends and colleagues, notably G. W. Bowersock, Pierre Briant, P. M. Fraser, the late J. F.
Gilliam, Chr. Habicht, and Javier Teixidor. It will readily be accepted that the remaining
imperfections are due to the author. There are many points at which systematic up-dating,
for instance as regards epigraphic finds, could be carried out. But that is a task for a new
study, and the chapter has been left as it was in .


The Hellenistic World and Rome
the Syrian region could have absorbed and reacted to the fact of Greek conquest. That this was true of the Jewish community of Jerusalem is beyond
all question.1  Maccabees, written originally in Hebrew, directly continues
the tradition of Old Testament historiography. It has indeed also been argued that Chronicles and the books of Ezra and Nehemiah were also written
in the Hellenistic period.2 If that is dubious, the book of Ecclesiasticus (Ben
Sira) was certainly written around  .. or soon after, and Daniel, in its
final form, in the s.3
The culture of Judaea and Jerusalem thus exhibits both a profound continuity with the pre-Greek past and an equally undeniable absorption of Greek
elements.4 As is well known, the first attested use of the word hellenismos
comes in  Maccabees (.) and refers to the enrolment of the Jerusalemites
as ‘‘Antiochians,’’ the setting up of a gymnasium, and the wearing of Greek
clothes.
We can therefore use Maccabees to pose at least one of the many questions
which can in principle be asked about Hellenistic Syria. By ‘‘Hellenistic’’ in
this sense I mean simply the period from Alexander to the mid-first century .. By ‘‘Syria’’ I mean anywhere west of the Euphrates and south of
the Amanus Mountains—essentially therefore the area west of the Euphrates
where Semitic languages were used: Aramaic in its various dialects, Phoenician, Hebrew, and earlier forms of Arabic. This begs a question about Asia
Minor (and especially Cilicia), from which Aramaic documents are known,
and a far more important one about northern Mesopotamia and about Babylonia. Should we not, that is, see the various Aramaic-speaking areas of the
Fertile Crescent as representing a single culture, or at any rate closely connected cultures, and therefore not attempt to study the one area without the
others?
The first question is one of cultural identity. Can we observe elsewhere in
Syria, that is, outside Judaea, either the continued survival of a non-Greek
culture or the fusion (Verschmelzung) in Droysen’s sense of Greek and nonGreek cultures? As I have argued elsewhere, there is perhaps just enough evi. F. Millar, ‘‘The Background to the Maccabean Revolution: Reflections on Martin
Hengel’s ‘Judaism and Hellenism,’ ’’ JJS  (): – ( chapter  in the present volume).
. O. Eissfeldt, The Old Testament: An Introduction (trans.) (), –.
. Schürer, Vermes, and Millar, History III, –, –. See also F. Millar, ‘‘Hellenistic History in a Near Eastern Perspective: The Book of Daniel,’’ in P. Cartledge, P. Garnsey,
and E. Gruen, eds., Hellenistic Constructs: Essays in Culture, History, and Historiography (),
– ( chapter  in the present volume).
. M. Hengel, Judaism and Hellenism: Studies in Their Encounter in Palestine during the Early
Hellenistic Period (trans.) I–II ().
The Problem of Hellenistic Syria

dence to show that this was the case in the Phoenician cities of the coast.5 But
elsewhere, with the exception of Judaea, we meet a problem which haunts
one and all of the questions we would like to ask. If we are going to ask
about the nature or limits of Hellenisation, there is a prior question: the Hellenisation of what? Whether we think of northern Syria, the Orontes valley,
or Damascus, or present-day Jordan, we find that almost nothing is known,
from either literary or documentary or archaeological evidence, about what
these places were like in the Achaemenid period.6 Our best evidence for the
personal life, nomenclature, and religious observances of non-Jewish Aramaic speakers in the Achaemenid period comes in fact from the private letters
in Aramaic from Egypt.7 The not very numerous monumental inscriptions
in Aramaic from Syria are no later than the seventh century ..8 The only
known cuneiform archive from Syria, found near Aleppo and dating to the
Neo-Babylonian and early Achaemenid periods,9 will serve to remind us of
how much we do not know. The only cuneiform tablet of the Achaemenid
period so far discovered in Jordan is, however, more revealing.10 Written in
Harran in the first year of a king named Darius, it records a sale by two people
with Aramaic names to a person whose father has the Edomite/Idumaean
name of Qusu-yada’. It was found at Tell Tawilan near Petra and thus clearly
reflects the type of movement and interchange round the Fertile Crescent
hinted at above. It is also significant that the same Idumaean name reappears
on an Aramaic-Greek bilingual ostrakon of the third century .. (text to n. 
below). By contrast, formal inscriptions in Aramaic are rare.11 Otherwise, it is
only in Teima in north-west Arabia, on the southern borders of what would
. F. Millar, ‘‘The Phoenician Cities: A Case-Study of Hellenisation,’’ Proc. Camb. Phil.
Soc.  (): – ( chapter  in the present volume).
. The archaeological evidence is largely confined to individual domestic or decorative objects and weaponry: see P. R. S. Moorey, ‘‘Iranian Troops at Deve Hüyük in the
Fifth Century ..,’’ Levant  (): –; P. R. S. Moorey, Cemeteries of the First Millenium
.. at Deve Hüyük (); only the evidence from the Judaean area has been systematically
assembled by E. Stern, The Material Culture of the Land of the Bible in the Persian Period ().
. J. T. Milik, ‘‘Les papyrus aramaéens d’Hermopolis et les cultes syro-phéniciens en
Egypte perse,’’ Biblica  (): ff.
. H. Donner and W. Röllig, Kanaanäische und aramäische Inschriften 2 I–III (–), nos.
–; A. Abou Assaf, P. Bordreuil, and A. R. Millard, La statue de Tell Fekherye et son inscription bilingue assyro-araméenne, Recherches sur les civilisations  (): Etudes assyriologiques ().
. F. M. Fales, ‘‘Remarks on the Neirab Texts,’’ OA  (): –.
. S. Dalley, ‘‘The Cuneiform Tablet from Tell Tawailan,’’ Levant  (): –.
. Though note that of Tobiah from ‘Araq el-Emir; see B. Mazar, ‘‘The Tobiads,’’ IEJ 
(): –, –.

The Hellenistic World and Rome
later be the Nabataean kingdom, that we can find Aramaic inscriptions, west
of the Euphrates and south of the Amanus, in the Achaemenid period itself.12
Aramaic ostraca of the Persian period are, however, known from a number
of sites in Israel, for example, Beer-Sheva and Arad.13 It can reasonably be
expected that archaeological investigation in areas outside present-day Israel
would produce more; and Aramaic material of the Persian period has, for
instance, been discovered at Tell el Mazar in Jordan.14
For the moment our evidence on Achaemenid Syria is very limited,15 and
what we know of its social and economic history is still largely dependent
on passing allusions in classical sources, for instance, Xenophon’s account of
his march across northern Syria from Myriandrus, a Phoenician trading post,
through an area of villages, and one satrapal palace and associated paradeisos,
to the city of Thapsacus on the Euphrates (Anab. ..–). There were apparently no cities on the route which they took between the coast and the
Euphrates at that moment. Did they deliberately avoid Aleppo, or had it declined as a city? Of the inland cities of the Syrian region which may still have
been inhabited in the Persian period, only Damascus is really certain. It was
there that Parmenio captured the treasures of Darius (Curt. ..; Arrian
Anab. ..–); and Strabo .. () says that it was the chief city of
Syria in the Persian period. Berossus also reports (FGrHist.  F ) that Artaxerxes II (/–/ ..) set up images of Artemis Anaitis in various
places, including Ecbatana, Babylon, Susa, Sardis, and Damascus.
Our ignorance of Achaemenid Syria is a major problem also for any assessment of the economic consequences of the Macedonian conquest. From
a ‘‘Marxist’’ standpoint, for instance, the late Heinz Kreissig argued that the
Seleucid empire continued to be based on the ‘‘Asiatic mode of production,’’ 16
. H. Donner and W. Röllig (n. ), nos. –; F. V. Winnett and W. L. Reed, Ancient
Records from North Arabia (); see G. Bawden et al., ‘‘The Archaeological Resources of Ancient Taymā: Preliminary Archaeological Investigations at Taymā,’’ Atlal  (): –;
new texts in A. Livingstone et al., ‘‘Taima: Recent Soundings and New Inscribed Material,’’
Atlal  (): ff.
. J. Naveh, ‘‘The Aramaic Ostracon,’’ in Y. Aharoni, ed., Beer-Sheba I: Excavations at Tel
Beer-Sheba – Seasons, Publications of the Institute of Archaeology  (), –;
J. Naveh, ‘‘The Aramaic Ostraca from Tell Arad,’’ in Y. Aharoni, Arad Inscriptions, Judaean
Desert Studies (), –.
. K. Yassine, ‘‘Tell el-Mazar, Field I. Preliminary Report of Area G. H. L. and M: The
Summit,’’ ADAJ  (): ff.
. See, e.g., A. F. Rainey, ‘‘The Satrapy beyond the River,’’ AJBA (): –.
. H. Kreissig, Wirtschaft und Gesellschaft im Seleukidenreich: Die Eigentums-und die Abhängigkeitsverhältnisse, Schriften zur Geschichte und Kultur der Antike  ().
The Problem of Hellenistic Syria

meaning the labour of peasants who were not slaves and owned their own
means of production but were dependent on those to whom they paid their
surplus. Pierre Briant, from a similar standpoint, once equated the ‘‘Asiatic
mode of production’’ with the ‘‘royal economy’’ briefly sketched in the Aristotelian Oeconomica .17 But if we look for specific and provable instances of
dependent villages in Syria in the Achaemenid period, we will find precisely,
and only, those in northern Syria which Xenophon states had been granted
to Parysatis (Anab. ..). We need not dispute Briant’s generalisation that the
village was a predominant social formation throughout the Near and Middle
East through both the Achaemenid and the Hellenistic periods. But we do
not know what was the typical set of existing economic relationships into
which the Macedonian conquest obtruded.
The fact of military conquest is indeed about all that is clear from the early
Hellenistic period. Beyond that we would want to ask, for instance, some of
the following questions: () What new Greek cities were founded, when, and
where? () Were they accompanied by Greek or Macedonian settlement in
the surrounding territories? () What substantial changes, if any, accompanied the acquisition of Greek names by existing cities? () Was there significant immigration and settlement by Greek speakers outside the context of
city foundations? () Are we to think of a degree of social and cultural fusion
between Greek settlers and the existing population, or rather, as Briant has
argued,18 of the Greeks forming separate enclaves? () Did the period see the
introduction into Syria of what ‘‘Marxists’’ define as the ‘‘ancient mode of
production,’’ that is, one based on a monetary economy, private property, and
the exploitation of slave labour? Any temptation to make sweeping generalisations in this topic should be tempered by the important evidence of the
papyri from the Wadi Dâliyeh, north of Jericho.19 They date to the third quarter of the fourth century, and may well have been deposited in the cave where
they were found in the aftermath of the Samaritan rising of circa  ..
One document of  .. records the sale of a slave for thirty-five pieces
of silver. There were also a number of coins, imported and local (especially
Tyrian), as well as seal impressions.
These documents are also potentially relevant to a final question: () What
. P. Briant, ‘‘Colonisation hellénistique et populations indigènes: la phase d’installation,’’ in Rois, tributs et paysans: études sur les formations tributaires du Moyen-Orient ancien, Centre
de Recherches d’Histoire Anciennes  (),  (originally published in Klio  []:
–).
. Briant (n. ) –.
. M. J. W. Leith, Wadi Daliyeh I: The Wadi Daliyeh Seal Impressions, DJD XXIV (),
and D. M. Gropp, Wadi Daliyeh II: The Samaria Papyri from Wadi Daliyeh, DJD XXVIII ().

The Hellenistic World and Rome
changes were brought about, outside the area of Greek settlement, in the
culture of the inhabitants, for example, in literacy? What combination of literacy was there in Semitic languages (Hebrew, Jewish Aramaic, Phoenician,
and later Nabataean Aramaic), in Greek, in both, or in neither?
The only substantial area where it is beyond question that new city foundations transformed the map of the region is northern Syria, with Seleucus I’s foundation of Antioch, Apamea, Seleucia, and Laodicea, a process brilliantly described by Seyrig.20 Near Antioch there was said to have been briefly
a city ‘‘Antigoneia,’’ founded by Antigonus Monophthalmus and settled by
Athenians (so Malalas, apparently following a chronographer named Pausanias, FGrHist.  F ). At Laodicea there was similarly said to have been
a village called ‘‘Mazabda,’’ and at Apamea one called ‘‘Pharnace’’ (FGrHist.
F , –).
Excavations on this site have revealed one object from the Persian period,
a fragment of an Attic pyxis.21 What is significant is that it is only, so far as I
have discovered to date, in the area of these cities that we find smaller settlements with Greek or Macedonian names. For instance, Diodotus Tryphon,
who seized the Seleucid throne in the s, came from a phrourion (fortified
settlement) called ‘‘Cassiana,’’ which, like others with the names ‘‘Larissa,’’
‘‘Megara,’’ ‘‘Apollonia,’’ and so forth, belonged to Apamea, where Tryphon
was educated (Strabo .. []). Even so, there were also villages in the
territory of Apamea with non-Greek names, like the kōmē Kaprozabadaiōn
(the village of the Karpazabadaioi) which an inscription reveals.22 The word
‘‘Kapro’’ reflects Kfar, meaning ‘‘village,’’ in Aramaic, as in Hebrew.23 Some
thirty miles east of Antioch, there was a village with the name ‘‘Maroneia,’’
which may be Greek; but, at any rate in the fourth century .., a person from there would speak Syriac ( Jerome, V. Malchi ). Similarly, twenty
kilometres north of Laodicea there was a place called ‘‘Heraclea Thalasse’’
(IGLS IV, , of / ..; cf. Pliny, NH , ). If there was any area where
Greek settlement may have produced significant direct effects on property
. H. Seyrig, ‘‘Antiquités Syriennes, : Séleucus I et la fondation de la monarchie
syrienne,’’ Syria  (): – slightly earlier version in English: ‘‘Seleucus I and the
Foundation of Hellenistic Syria,’’ in W. A. Ward, ed., The Role of the Phoenicians in the Interaction of Mediterranean Civilization, Papers Presented to the Archaeological Symposium at the
American University of Beirut, March  (), –.
. J. Balty and J. C. Balty, ‘‘Apamée de Syrie, archéologie et histoire I. Des origines à
la Tetrarchie,’’ ANRW II. (), –.
. IG XIV ; photo in J. C. Balty, ‘‘Nouvelles données topographiques et chronologiques à Apamée de Syrie,’’ AAAS  (): –; see BE , .
. See BE , no. .
The Problem of Hellenistic Syria

relations and ‘‘modes of production,’’ it will have been in the territories surrounding the new cities of the north Syrian tetrapolis.
Elsewhere concrete evidence for new city foundations of the earlier Hellenistic period is remarkably sparse. There were none along the Phoenician
coast or in Idumaea or Judaea. Late sources record that a Macedonian settlement was established by Alexander or Perdiccas on the site of Samaria.24 In
this case there is substantial archaeological evidence which can be brought
into relation with this settlement. The round towers added to the existing
wall of the acropolis are dated to the late fourth century; an outer circuit
of walls, with square towers, perhaps belongs to the second century ..25 It
seems certain that this small fortified town on a hill-top is that of the Macedonian settlers and their descendants, to be distinguished from the Samaritans proper, who in the later fourth century had established their own temple
on Mount Gerizim.26
The cults followed by the settlers are illustrated (if no more than that) by
a finely cut inscription of the third century .. from Samaria with a dedication by Hegesander, Xenarchis, and their children to Sarapis and Isis.27 But
in many places we cannot be certain what social changes are implied by the
appearance of cities with Macedonian place-names, like Beroea, Cyrrhus,
or Gindarus in the north-east, or Pella or Dium in Jordan.28 Excavations at
Pella have revealed some evidence of the earlier Hellenistic period.29 But
Cyrrhus, for example, makes no appearance at all in our sources until ,
‘‘Cyrrhestian’’ soldiers are recorded as mutinying against Antiochus III in
.30 It is reasonable to believe that it was a Macedonian settlement of the
early period, like Dura-Europos on the Euphrates. But again, very little is
known of the social character of Hellenistic Dura except the vital item that
. Schürer, Vermes, and Millar, History II, .
. J. W. Crowfoot, K. M. Kenyon, and E. L. Sukenik, Samaria-Sebaste I: The Buildings
(), –.
. Schürer, Vermes, and Millar, History II, ; see further text to n.  below.
. J. W. Crowfoot, G. M. Crowfoot, and K. M. Kenyon, Samaria-Sebaste I: The Objects
(London, ), , no. ; illustrated in M. Avi-Yonah, The Holy Land (), .
. For toponyms in Syria, see the illuminating survey by E. Frézouls, ‘‘La toponymie
de l’Orient syrien et l’apport des éléments macédoniens,’’ in La toponymie antique: actes de
colloque de Strassbourg  (), –.
. R. H. Smith, ‘‘Preliminary Report of the  Season of the Sydney/Wooster Joint
Expedition to Pella,’’ ADAJ  (): ff.
. Polyb. , , –; , ; see E. Frézouls, ‘‘Recherches historiques et archéologiques sur
la ville de Cyrrhus,’’ AAAS – (–): –; E. Frézouls, ‘‘Cyrrhus et la Cyrrhestique
jusqu’a la fin du Haut-Empire,’’ ANRW II. (), –.

The Hellenistic World and Rome
some at least of the land there was classified as klēroi (lots; P. Dura , second
century ..). As a physical structure it was, like Samaria, a fortified site of
moderate extent (the three longer sides measuring just under , metres
each), sited on a plateau above the Euphrates, and later equipped with walls.
Internally, it was set out in regular blocks on the well-known Hippodamian
plan, with a central agora. It is uncertain which, if any, temples can be attributed to the initial Hellenistic phase. No evidence for a theatre or other public
buildings of this period has been found.31 It is natural to presume that we
should envisage both Samaria and Dura as Macedonian military settlements,
placed for strategic purposes in alien landscapes, and with modest pretensions to being the bearers of a wider Greek culture. In the case of many other
foundations there is still less evidence. Beroea (Aleppo) is recorded as a foundation of Seleucus I (Appian, Syr. ). Once again we have nothing to show
whether the ancient city of Aleppo still existed at the moment of the settlement; but the street plan to this day reflects the rectangular axes which may
well be those of the colony.32
The same problems persist if we look at places which subsequently gained
Hellenistic dynastic names: Philadelphia (Amman) and Ptolemais (Acco)
from Ptolemy Philadelphus (near here Strabo .. [], notes three placenames which may reflect Ptolemaic rule: ‘‘Sykaminōn polis,’’ ‘‘Boukolōn
polis,’’ and ‘‘Krokodeilōn polis’’). Epiphaneia (Hama) presumably gained its
name from Antiochus IV Epiphanes. This was of course another ancient city,
which, as Josephus records, the natives (epichōrioi) still called ‘‘Hama’’ (Ant. ,
). But, paradoxically, excavations on the site have seemed to suggest that
it was unoccupied between its destruction by Sargon II in  and the beginning of Greek settlement in the second century ..33 On the other hand,
Sargon is recorded to have settled , Assyrians there, and there continue
to be occasional mentions of Hama as a place in documents of the intervening period;34 the archaeological evidence should not be interpreted on
the assumption that the site was desolate after ,35 and imported Hellenistic
. A. Perkins, The Art of Dura-Europos (), –; see F. Millar, ‘‘Dura-Europos under
Parthian Rule,’’ in J. Wiesehöfer, ed., Das Partherreich und seine Zeugnisse, Historia-Einzelschrift  (), – ( chapter  in the present volume).
. J. Sauvaget, Alep: essai sur le développement d’une grande ville syrienne des origines au milieu
du XIX e siècle (), .
. So E. Fugman, Hama. Fouilles et recherches de la Fondation Carlsberg – II: architecture des périodes pré-hellénistique (), .
. Fugman (n. ), .
. See esp. E. D. Francis and M. Vickers, ‘‘Greek Geometric Pottery at Hama and Its
Implications for Near Eastern Chronology,’’ Levant  (): ff.
The Problem of Hellenistic Syria

pottery appears there before the reign of Antiochus IV.36 The evidence for
continuity of settlement is therefore ambivalent; and while the evidence for
the Hellenistic city remains unpublished, it is impossible to say whether it
would suggest the implantation of an organised settlement at a specific moment. But if Epiphaneia did receive an actual settlement of Greeks, there
was certainly no such settlement in Jerusalem in the s, when the population briefly acquired the title ‘‘Antiocheis.’’ The settlement on the Akra in
Jerusalem in the s was another matter.
The provable extent of organised Macedonian or Greek settlement is thus
limited to one area, north Syria. Other towns which acquired Greek names
may well also have received settlements, but some certainly did not. If we
consider the entire non-desert area west of the Euphrates, Greek colonial
settlement must be regarded as a relatively limited phenomenon, largely restricted in time also, to the reign of Seleucus I. Whatever created the conditions for a large-scale transformation, fusion, or conflict, if anything did,
it was not, except in northern Syria, a massive process of colonisation.
Was there none the less extensive private immigration, either for settlement on the land or for other purposes, such as trading in slaves? Here again
we have to say that we do not know. We can easily illustrate, for instance, the
presence in Syria of Ptolemaic soldiers from various parts of the Greek world;
the inscription from Ras Ibn Hani on the coast eight kilometres north of
Laodicea 37 which records some of these is the earliest Greek public inscription from northern Syria, dating to about the second half of the third century. Excavations on this site have shown that a fortified Greek town, whose
name remains unknown, was established there in the same period, probably
by the Ptolemies.38 Greeks also entered the service of local dynasts: a papyrus
from the Zenon archive shows us soldiers from Cnidus, Caunus, Macedon,
Miletus, Athens, and Aspendus serving in  under Tobias in Ammonitis
(PCZ  CPJ I, ). In the second century we come across a Macedonian
settled at Abae in Arabia and married to an Arabian wife (Diod. Sic. , , ),
or a politeuma of Caunians settled in Sidon (OGIS ). No doubt we could
accumulate further illustrations; but it would hardly be significant, since it
would be more than surprising if there had been no Greek private settlement
in this region. But it does have to be emphasised that there is no positive evi. A. P. Christensen and C. F. Johansen, Hama. Fouilles et recherches – III.: les
potéries hellénistiques et les terres sigillées orientales (), .
. J.-P. Rey-Coquais, ‘‘Inscription grecque découverte à Ras Ibn Hani: stèle de mercénaires lagides sur la côte syrienne,’’ Syria  (): –.
. P. Leriche, ‘‘La fouille de la ville hellénistique d’Ibn Hani: bilan provisoire ,’’ in
M. Yon, ed., Archéologie au Levant: recueil à la mémoire de Roger Saidah (), ff.

The Hellenistic World and Rome
dence to suggest that there was private immigration on a scale which would
by itself have brought profound changes in culture, social relations, or the
economy.
If we go back to the major cities of the Syrian tetrapolis, there is certainly adequate evidence to illustrate their character as Greek cities in the
Hellenistic period. It should be stressed that in the absence of large-scale
documentary evidence we still depend quite significantly on passing items
of narrative material, like the papyrus report (the Gurob papyrus) from the
Ptolemaic side, of Ptolemy III’s invasion of Syria in the s.39 It records the
priests, archontes (annual magistrates), and the other citizens of Seleucia, with
the hēgemones (officers of the Seleucid garrison) and soldiers, coming down to
greet the invading forces. A similar scene is said to have followed at Antioch,
with a ceremonial greeting before the city by ‘‘satraps and other hēgemones,
and soldiers and synarchiai [magistrates] and all the neaniskoi [youth] from the
gymnasium,’’ and the rest of the population, bearing cult images. Seleucia
does not reappear in our evidence until we come to Polybius book , and the
narrative of its recapture by Antiochus III in ; it turns out to be a place
of modest size, with some , ‘‘free men’’ (which may mean citizens only,
or all non-slave male inhabitants); perhaps therefore some , persons in
all (, , ). These cities produce nothing like the vast harvest of monumental Greek inscriptions which characterise (say) Delphi, Delos, or some of the
Greek cities of western Asia Minor in this period. It is true that at Antioch,
Seleucia, and Laodicea subsequent occupation greatly limits the possibilities
of excavation; there is some scanty evidence on early Hellenistic Laodicea.40
But it should still not be assumed that the social conditions which elsewhere
led to the large production of public inscriptions necessarily applied in Syria
in the same way. Public inscriptions from Seleucia in Pieria do reveal, for
instance, the vote of a statue for the Seleucid epistatēs (overseer) of the city
in  ..;41 or a letter of Antiochus VIII or IX, of  .., confirming the
freedom of the city.42 There is no substantial corpus of the public inscriptions of Seleucia; excavation of the relevant public buildings, when identified, might of course reveal them. From Laodicea the only known public decision recorded on stone from the Hellenistic period is the gnōmē (proposal)
. FGrHist. ; see translation in M. M. Austin, The Hellenistic World from Alexander to
the Roman Conquest: A Selection of Ancient Sources in Translation (), no. .
. See R. A. Stucky, Ras Shamra—Leukos Limen: Die nach-ugaritische Besiedlung von Ras
Shamra (), –.
. C. B. Welles, Royal Correspondence in the Hellenistic Age (New Haven, , repr. ),
no.  IGLS ., .
. OGIS ; Welles (n. ), nos. –.
The Problem of Hellenistic Syria

of Asclepiades and the archontes, approved by the peliganes (the councillors, a
Macedonian word) in  .. (IGLS IV, ), concerning the sanctuary of
Isis and Sarapis. From Antioch and Apamea there are no public decrees at all
surviving from the Hellenistic period; though one inscription from Antioch shows theōroi (sacred delegates) honouring an agōnothetēs (director of an
agōn, i.e., public games) from Seleucia in / ..;43 and one from Daphne
shows Antiochus III appointing a priest there.44 Passing literary references
indicate at least the existence of gymnasia at Laodicea (Appian, Syr. ) and
at Daphne near Antioch (Polyb. , , ) and, for example, of a hippodrome
near Seleucia (Polyb. , , ). Posidonius’ remarks on the luxury of life in
Syria (Ath. e–f e–f ) imply that gymnasia were common. None of
these cities, however, has revealed any trace of a theatre that can be firmly
dated to this period. It is surely, I think, a revealing fact that there is no certain archaeological evidence for a theatre of the Hellenistic period anywhere
in the Syrian region. Given the relative indestructability of theatres built
against hillsides, as Hellenistic theatres normally were (e.g., those of Priene
or Delos), this is one case where negative evidence may be suggestive.45
Outside the places which we know to have been royal foundations, or to
have acquired Greek names, we do have some evidence, from various periods, of the spread of a recognisably Greek way of life. A site called Ayin Dara,
north-east of Aleppo, for instance, shows traces of occupation in the Persian period and then a substantial urban area with walls from the Hellenistic
period, with pottery and coins of the second and first centuries ..46 This
site, whose Greek name, if it had one, is unknown, is a reminder of just how
much of the evidence of Hellenistic Syria may simply be lost. For contrast
we have Tel Anafa in northern Galilee, whose heated bath-house of the later
second century .. is the earliest known from the Near East;47 and the wellknown site of Marisa in Idumaean, a small urban settlement of six acres, built
in the third or early second centuries .., and enclosed by a fortification
wall. Greek was in use there, as shown by some execration texts in Greek,
. C. H. Kraeling, ‘‘A New Greek Inscription from Antioch on the Orontes,’’ AJA 
(): ff. (and pl. ); BE , .
. Welles (n. ), no. .
. So E. Frézouls, ‘‘Recherches sur les théatres de l’Orient syrien,’’ Syria  (): –
; but see M. A. R. Colledge, ‘‘Greek and Non-Greek Interaction in the Art and Architecture of the Hellenistic East,’’ in A. Kuhrt and S. M. Sherwin-White, eds., Hellenism in the
East (), .
. E. Seirafi and A. Kirichian, ‘‘Recherches archéologiques à Ayin Dara au N-O d’Alep,’’
AAAS . (): –.
. S. Herbert, ‘‘Tel Anafa: The  Season,’’ Muse  (): ff.

The Hellenistic World and Rome
and the inscriptions on the well-known painted tombs. But the house-types
are non-Greek and at least some of the inhabitants identified themselves in
Greek as ‘‘Sidonians in Marisa.’’ 48 The mixed culture of this area in the third
century .. is vividly illustrated by a group of ostraca from Khirbet el-Kōm,
four in Aramaic, one in Greek, and one Greek-Aramaic bilingual; the latter
records the borrowing of thirty-two ZUZN by Niceratus from one Qusyada’/Kosides, described (in both texts) by the Greek word kapēlos, ‘‘trader.’’ 49
This text, probably of  .., thus reveals kapēlos as a loan-word in a dialect of Aramaic. These ostraca are closely paralleled by an Aramaic ostracon
of the third century from Jerusalem, also containing what seem to be two
Greek loan-words.50
The ostracon is given the date  on the supposition that the ‘‘sixth year’’
referred to in it is that of Ptolemy Philadelphus. Most of the evidence which
illustrates Greek economic activity in Syria comes from the Ptolemaic phase
of control. All we have is fragments, occasionally illuminating. Some cities,
as we saw, gained dynastic names, Akko becoming ‘‘Ptolemais,’’ and RabbatAmman ‘‘Philadelphia.’’ Scythopolis and Philoteria in Galilee must also have
gained their Greek names in the Ptolemaic period. Our main evidence comes
from the Zenon papyri, discussed by Tcherikover.51 These papyri, of course,
owe their survival to the particular conditions of Egypt, and thus cast a sidelight on Syria somewhat comparable to that shed by the Aramaic documents
from Egypt of the Achaemenid period (text to note  above). But while
the relevant climatic and soil conditions, allowing the survival of perishable
writing materials, very rarely apply in the Near East, they are not wholly
unknown, as the documents from the Judaean Desert show.
The first thing that the Zenon papyri clearly illustrate is the slave trade.
Among the immigrant Greek klērouchoi (colonists) serving at Birtha in Ammonitis under Tobias, mentioned earlier (PCZ  CPJ I, ), one sells
a slave girl named Sphragis, apparently from Babylon or Sidon, to another,
who then sells her to Zenon. In Marisa Zenon also bought some slaves (sōmata), two of whom escaped and had to be searched for (). One Menecles appears as having transported some slaves and other merchandise ( phorta)
from Gaza to Tyre, and as intending to tranship them without paying the ex. OGIS ; G. Horowitz, ‘‘Town Planning in Hellenistic Marisa: Reappraisal of the
Excavations after Eighty Years,’’ PEQ  (): –.
. L. T. Geraty, ‘‘The Khirbet el-Kōm Bilingual Ostracon,’’ BASOR  (): –.
. F. M. Cross, ‘‘An Aramaic Ostracon of the Third Century .. from Jerusalem,’’ EretzIsrael  (): ff.
. V. Tcherikover, ‘‘Palestine under the Ptolemies (a Contribution to the Study of the
Zenon Papyri),’’ Mizraim – (): –.
The Problem of Hellenistic Syria

port tax or having an export permit (exagōgē tōn sōmatōn) (PCZ ). Tobias
also sends to Apollonius a group of four slaves as a gift, two described as circumcised and two not (PCZ  CPJ I, ). There is no obvious reason in
the text for regarding either of the circumcised men as Jewish; if they were
not, then this is evidence for the continuation of the custom of circumcision
among the Syrians generally in the Hellenistic period.
Much more informative for the continuity of a non-Greek culture is a
papyrus letter of  .. from Egypt mentioning a slave who was ‘‘by race a
Syrian from Bambyce’’ who was ‘‘tattooed on the right wrist with two barbarian letters.’’ 52 The letters can only have been Aramaic ones; Bambyce is
Hierapolis, an important centre of a non-Greek culture, on which see further below. Lucian, On the Syrian Goddess, written in the second century ..,
which is about the goddess Astarte of Hierapolis, records () that the Syrian
adherents of the cult were tattooed on the neck or wrist—‘‘and thence all
Assyrioi bear tattoos.’’
To come back to the economic impact of Greek rule, slavery and the slave
trade were clearly a feature of it; but whether this was a novelty remains unclear. The most striking reflection of slavery in Hellenistic Syria remains the
well-known edict of Ptolemy II Philadelphus dating to April , which
surely can be taken to illustrate a causal connection between foreign domination and slavery:53 ‘‘If any of those in Syria or Phoenicia have bought a
sōma laikon eleutheron [free native],’’ or have acquired one in any other way,
they are required to prove that they were slaves at the time of acquisition.
Those bought at royal auctions, however, are legally owned even if they
claim to be free. Moreover soldiers or others who are settled (katoikountōn)
who are living with gunaikes laikai (native wives) need not declare them as
slaves. In the future it will be forbidden to acquire possession of sōmata laika
eleuthera except those sold up by the superintendent of the revenues of Syria
and Phoenicia. Whatever the legal definitions involved, the order clearly reflects a notion of the particular liability of the free ‘‘native’’ population of
south Syria to slavery; in particular it is significant that the document has
to make clear that the ‘‘native’’ wives of Ptolemaic soldiers and settlers do
not have to be categorised as slaves. Strikingly, in this case, the king is taking
steps to limit the circumstances under which individuals found themselves
regarded as slaves.
. See G. Vaggi, ‘‘Siria e Siri nei documenti dell’Egitto greco-romano,’’ Aegyptus 
(): –, on –.
. M.-T. Lenger, Corpus des ordonnances des Ptolémées 2 (Brussels, ), no. ; Austin
(n. ), no. .

The Hellenistic World and Rome
As was hinted earlier, any notions of what social, economic, cultural, or
social status is implied by the expression sōmata laika eleuthera must remain
hypothetical. Even if we disregard acute regional variations (see below), it is
no use pretending that we have any idea of the typical forms of property relations in the Syrian area either before or after Alexander’s conquest. We can
of course see examples of various different things, for instance, the exploitation of private landed property in the Ptolemaic period in Palestine, perfectly
exemplified in a papyrus from the Zenon archive (P. Lond. ) of  ..
An agent, Glaucias, is writing to Apollonius about his enormous vineyard
at Bethaneth, which was somewhere in Galilee: ‘‘On arrival at Bethaneth I
took Melas with me and inspected the plants and everything else. The estate
seems to me to be satisfactorily cultivated, and he said the vines numbered
,. He has also constructed a well, and satisfactory living quarters. He
gave me a taste of the wine, and I was unable to distinguish whether it was
Chian or local. So your affairs are prospering and fortune is favouring you all
along the line.’’ This does on the face of it seem to be an example of the deliberate increase of productive capacity of a sort which, for Ptolemaic Egypt
generally, Alan Samuel sought to deny.54 There is, however, no indication
of how the estate was worked, whether by slave labour, free hired labour,
or dependent villagers. The question of dependent, but non-slave, agricultural labour in the Hellenistic world has attracted an enormous amount of
attention. But the evidence comes almost entirely from Seleucid land grants
or sales in Asia Minor; moreover, the real social and economic relations alluded to in these inscriptions remain extremely obscure.55 It is also from
Asia Minor, and entirely from Strabo’s Geography, that we have almost all
the available descriptions of large communities of hierodouloi (sacred slaves)
attached to temples.56 Comparable evidence is hardly available from Syria.
There are, I think, just three items. First is the mutilated inscription from
Hephzibah near Scythopolis (Bethshean) first published by Landau  and
re-edited by Fischer .57 The dossier contains petitions to Antiochus III
from Ptolemaeus, described as stratēgos (general) and archiereus (high priest),
. A. E. Samuel, From Athens to Alexandria: Hellenism and Social Goals in Ptolemaic Egypt
(Louvain, ).
. Welles (n. ), nos. –, –, ; cf. P. Briant, ‘‘Remarques sur ‘laoi’ et esclaves
ruraux en Asie Mineure hellénistique,’’ in Rois, tributs et paysans (n. ), – (originally
published in Actes du colloque  sur l’esclavage []: –).
. K. W. Welwei, ‘‘Abhängige Landbevölkerung auf ‘Tempelterritorien’ im hellenistischen Kleinasien und Syrien,’’ Ancient Society  (): –; P. Debord, Aspects sociaux et
économiques de la vie religieuse dans l’Anatolie gréco-romaine (Leiden, ), ff.
. Cf. J. M. Bertrand, ‘‘Sur l’inscription d’Hefzibah,’’ ZPE  (): –.
The Problem of Hellenistic Syria

and the king’s replies, concerning properties owned by Ptolemaeus. The context is immediately after the Seleucid conquest in . All that is clear is that
the dossier speaks of kōmai (villages) ‘‘of ’’ Ptolemaeus, and that as owner he is
concerned to protect the people (laoi ) in his villages from official exactions
and the quartering of troops (epistathmeia). One phrase may imply that some
kōmai had been inherited by him, and others added by the king’s command.
We can accept that the document embodies the notion of owning villages
and (in some sense) of owning or being responsible for the people who inhabit them.
Further north, from the hills inland from Aradus, we have the famous inscription of the temple of Baetocaece.58 In response to a report about the
energeia (power) of the god Zeus of Baetocaece, King Antiochus—which one
is uncertain 59—announces his decision to grant to the god the village of
Baetocaece, which a certain Demetrius formerly ‘‘had’’ (eschen), so that its
revenue ( prosodos) may be spent on the sacrifice, and any other steps taken
for the improvement of the shrine by the priest appointed by the god. There
is to be a monthly tax-free fair, the sanctuary is to have the right of asylum,
and the village is on no account to be subject to billetting.
It is clear that the cult of Zeus of Baetocaece already existed. The village
had, up to the moment of the king’s grant of it to the temple, been in private
possession. This may mean no more than that there had previously been a
(revocable) grant of it to a private person by an earlier king; that is to say that
the village belonged in the category of chōra basilikē, royal land.60 No such
legal prescription is actually stated in the document, and it is clear from the
king’s decision that some representation had been made to him about the
‘‘power’’ (energeia) of the god. It is, therefore, equally possible that he is approving the transfer to the sanctuary of land which had previously been in
full private ownership. Exactly what is meant for the status of the inhabitants is uncertain. In inscribing this document in the s .., and also a
little earlier, in the s (IGLS VII ), they describe themselves as katochoi
(subjects) of the god.
The city of Aradus is not involved in this initial transaction, though it was
later, under Augustus. Are we then dealing with royal land (chōra basilikē )
either in the sense of an individual royal property or in the wider sense, com. IGLS VII, ; Austin (n. ), no. .
. H. Seyrig, ‘‘Antiquités syriennes, : Arados et Baetocaece,’’ Syria  (): –;
K. J. Rigsby, ‘‘Seleucid Notes,’’ TAPhA  (): –; B. Baroni, ‘‘I terreni e i privilegi
del tempio di Zeus a Baitokaike (IGLS VII ),’’ in B. Virgilio, ed., Studi ellenistici I (),
–.
. So Baroni (n. ).

The Hellenistic World and Rome
monly imagined in modern books, that all the land outside city territories
was ‘‘royal,’’ that is, in some sense owned and exploited by the king, and at
his disposition? In my view this notion goes far beyond what our evidence
shows. As what is said below will illustrate, it is very questionable whether
this concept has any reflection in the real-life social and economic relations
which our sources attest. There were many non-city areas where no direct
control was, or could be, exercised by any king.
Where we do find land in royal possession, and then being assigned for
cult purposes, is in the remarkable documents from Commagene in which
Antiochus I (c. / ..) proclaims his institution of a cult for various gods,
for his deceased father, Mithridates Callinicus, and for himself.61 Among the
other provisions he states that he has dedicated a group of mousikoi who
are to learn the arts necessary for performing at the cult festivals, and to be
succeeded in the same skills by their sons, daughters, and all their descendants. They are described as hierodouloi (sacred slaves), and are to maintain
this hereditary role forever. It is not clear, however, whether these are or are
not the same as the inhabitants of the kōmai (villages) which, in the Nemrud
Dagh text, he says he had dedicated to the gods, or (in the text from Arsamea
in the Nymphaios) of the land ek basilikōs ktēseōs (in royal possession) which
he has dedicated with its revenues, to be looked after by the priests. But at
least we confront here an unambiguous reference to specific royal properties,
and also, once again, a category of non-free persons (hierodouloi ) which does
not descend from a remote past but is being created in the first century ..
Not far away, and at about the same time, Cicero fought his miserable
little campaign against the free Cilicians of the Amanus, whose town, Pindenissum, was high up, well fortified, and inhabited by people who had never
yielded obedience to the (Seleucid) kings ( fam. , , ). It took him a siege
of fifty-six days to capture it. The mountainous or marginal areas of the
Syrian region were covered with fortified villages, whose inhabitants, as far
as we can see, were integrated in no system of property relations imposed
from outside and did not belong in any functional sense to any state. Internally, of course, they had their own systems of social stratification. We see
this best in one vivid report which relates to two village communities in
Moab in about  .. A people called the ‘‘sons of Jamri’’ were celebrating
the wedding of the daughter of one of the notables (megistanes) of Canaan,
conducting the bride in a great procession laden with possessions. From the
. H. Waldmann, Die kommagenischen Kultreformen unter König Mithridates I Kallinikos und
seinem Sohne Antiochos I (); J. Wagner and G. Petzl, ‘‘Eine neue Temenos-Stele des Königs
Antiochos I von Kommagene,’’ ZPE  (): –.
The Problem of Hellenistic Syria

opposite direction the bridegroom with his friends and brothers was on his
way to meet the bride, accompanied by musicians playing tambourines, and
an armed escort. At that point the scene stops; for Jonathan and Simon Maccabaeus with their followers leap up from ambush, slaughter as many as they
can, put the rest to flight, and take all their possessions ( Macc. :–).
The two books of Maccabees, especially the first, give us the best—and
more or less contemporary—picture which we have of social formations
and settlement patterns in the southern part of the Syrian region in the second century; they would deserve further investigation, directed to the hints
which they provide as to non-Jewish social structures in this period. The
Maccabean wars stretched from the cities of the Philistine coast, like Azotus
with its temple of the Philistine god, Dagon (‘‘Bethdagon,’’  Macc. :–
), to the fortified villages (ochuromata) of Idumaea ( Macc. :) or Transjordan ( Macc. :–). In  Macc. :– a whole string of places across
the Jordan, all of which have retained analogous Arabic names until modern times—‘‘Bosora,’’ ‘‘Bosor,’’ ‘‘Alema,’’ ‘‘Chaspho,’’ ‘‘Maked,’’ ‘‘Karnaim’’—
are described as large, fortified poleis. These too will have been fortified villages; it is worth noting that the author of  Maccabees has no notion that
term polis ought to be restricted to self-governing cities formally recognised
as such; he uses it for instance of Modein (:), the village from which the
Maccabees came.62 Similarly, Polybius uses the word polis of Atabyrion, a
settlement on Mount Tabor (, , ).
The narratives of Maccabees also illustrate the very close geographical
conjunction between different social or economic groupings which characterised this area, since the operations bring the Jewish forces into repeated
contact not only with cities and with fortified villages, but with groups described as ‘‘Arabs,’’ following a nomadic, or at any rate non-sedentary, way
of life. Even on the coastal strip near Jaffa, Judas Maccabaeus is attacked by
no less than , Arabs with  horsemen, described as nomades. When
defeated, they offer cattle as a pledge of friendship and retire to their tents
(skēnai;  Macc. :–). The social pattern of an intermingling and mutual
dependence, balanced by recurrent hostilities, between various gradations of
settled, pastoral, and truly nomadic communities using camels, is of course
well known, and nowhere better described than by Donner on the early
Islamic conquests.63 It is worth noting that Diodorus, concluding his account
of the Nabataeans, gives a succinct account of the social relations involved
. But see R. J. van der Spek, ‘‘The Babylonian City,’’ in Kuhrt and Sherwin-White
(n. ), .
. F. M. Donner, The Early Islamic Conquest ().

The Hellenistic World and Rome
(, , ): ‘‘There are also other tribes of Arabs, of whom some even cultivate the soil, intermingled with the tax-paying peoples, and [who] share the
customs of the Syrians, except that they dwell in tents.’’
I return below to the question of the movement of Arab peoples into
Syria and their settlement there, a subject discussed in an interesting way by
Dussaud.64 For the moment note Diodorus’ contrast between Arabs living in
tents and those in settled populations who can be made to pay taxes. In many
parts of the Syrian region, in the mountains and on the fringes of the desert
above all, the Seleucid (or Ptolemaic) state either never had or only occasionally had any effective presence as the Achaemenids before them, who,
however, maintained a contractual relationship with them.65
Most of what I have been saying so far has been designed to suggest how
limited, variable, and erratic the Greek presence in the different parts of the
Syrian region was in the Hellenistic period, at any rate so far as our present
evidence shows. I now wish to look at the other side, and ask what if anything
we know of the non-Greek culture of the area, or of essentially non-Greek
communities within it. The Phoenician cities of the coast preserved their
historical identity and culture, while evolving, by steps which we cannot
really trace, into Greek cities.66 A similar evolution seems to have taken place
in the ancient Philistine cities further south, Azotus, Ascalon, and Gaza.67 As
with many other places in the Near East, their non-Greek, or not wholly
Greek, identity is expressed most clearly in dedications made on Delos. Perhaps the most striking example is the well-known dedication by a man from
Ascalon: ‘‘To Zeus Ourios and Astarte Palestine and Aphrodite Ourania, the
listening gods, Damon son of Demetrios, of Askalon, having been saved from
pirates, [offers his] prayer. It is not permitted to introduce [here] a goat, pig,
or cow.’’ 68 The notion that these, or any other existing communities, could
be made into Greek cities purely by the issue of some sort of charter or the
granting of a Greek constitution, without either a settlement or building
operations, still seems to me to need further examination. It is more in accordance with the evidence to see these coastal cities as places which had
. R. Dussaud, La pénétration des Arabes en Syrie avant l’Islam (Paris, ); see also
P. Briant, Etat et pasteurs au Moyen-Orient ancien (), chap.  (an important study); and
I. Shahid, Rome and the Arabs ().
. Briant (n. ), ff.
. Millar (n. ).
. K. Rappaport, ‘‘Gaza and Ascalon in the Persian and Hellenistic Periods in Relation
to Their Coins,’’ IEJ  (): ff.; Schürer, Vermes, and Millar, History II, –.
. ID no. ; P. Bruneau, Recherches sur les cultes de Délos à l’epoque hellénistique et à
l’epoque impériale (), –.
The Problem of Hellenistic Syria

been in close contact with the Greek world before Alexander, and where,
after the conquest, a continued process of Hellenisation took place gradually against a background of cultural continuity.69 But we should not think
of the non-Greek elements as being static features of a world in which cultural change came only from the Greek side. For instance, some nine kilometres from Acco/Ptolemais a Greek inscription of probably the second century .. shows a man with a Greek name dedicating an altar to Hadad and
Atargatis, ‘‘the listening gods.’’ 70 Rather than being an example of the continuity of local non-Greek cults, this inscription is the earliest attestation of
these deities on the Phoenician coast.
A much greater problem is presented by those inland cities or communities which are not known to have received any formal Greek colony or
settlement. With Jerusalem and Judaea the essential features of cultural and
religious contact and conflict are well known: a significant level of Hellenisation, met by a conscious and violent reaction and reassertion of ‘‘national’’
tradition. The Samaritans too retained and reasserted their ‘‘Israelite’’ identity. This fact is perfectly illustrated by two dedications of the middle and
late second century from Delos in Greek put up by ‘‘The Israelites on Delos
who pay their tithes to holy Argarizein [Mount Gerizim].’’ 71
By contrast, if we think for instance of Damascus, virtually nothing is
known of its character as a city or community at the moment of the Macedonian conquest except the bare fact of its existence. Nor has any significant evidence about it through the Hellenistic period survived, beyond some
coins of the second and first centuries .. with the legend Damaskēnon (of
the people of Damascus) and passing mentions of it as an object of successive dynastic conflicts.72 The occasional documents of persons from Damascus abroad in the Hellenistic world are not very informative, though
they do illustrate the adoption of Greek nomenclature. It is not surprising
that Semitic names might also be retained, for instance ‘‘Martha, Damascēnē
[a Damascene],’’ on a late second-century inscription from Delos (ID nos.
–).
No real insight into the internal life of Damascus can be attained until the
middle of the first century, when we come to Nicolaus’ account of his father
Antipater, who was presumably born around the beginning of the century
. See Colledge (n. ), .
. M. Avi-Yonah, ‘‘Syrian Gods at Ptolemais-Accho,’’ IEJ  (): –.
. P. Bruneau, ‘‘ ‘Les Israélites de Délos’ et la juiverie délienne,’’ BCH  (): –
.
. Schürer, Vermes, and Millar, History II, –.

The Hellenistic World and Rome
and was a skilled orator (in Greek, as is clearly implied), who filled all the
offices (archai) in the city and represented it before the various dynasts who
ruled in the surrounding area (FGrHist.  F ). In the chaotic conditions
of fluctuating empires and local tyrannies which marked the history of Syria
in the middle of the century,73 this will have been an essential function. The
education and culture to which Nicolaus laid claim (FGrHist.  F ) was
wholly Greek, and nothing in the extensive fragments of his works suggests
any influence from a different historical or cultural tradition. In this he offers
an obvious and striking contrast to Josephus, who was to make so much use
of him as a source.74
A combination of different cultural traditions is certainly expressed in the
monuments and inscriptions of one local dynasty which emerged in northern Syria in the second century, the royal house of Commagene.75 But if
what we are interested in is a local ‘‘mixed’’ culture, Commagene is not a true
exception, for everything that we can observe there is, firstly, a royal invention; and, secondly, though the kings consciously draw on two traditions,
they do so in relation to Greek and Persian elements, not Syrian or Aramaic
ones: Greek gods and Ahuramazda; royal descent from Persia and Macedon;
Persian dress to be worn at festivals.76 It was natural, in creating a new royal
ideology, to look to the two major monarchies of the Achaemenids and the
Seleucids. But there is still a contrast, for instance, with the contemporary
coinage of the Hasmoneans in Judaea, which incorporates Greek, Hebrew,
and Aramaic elements.77
So far as I can find, a real continuity is traceable in just one place outside Phoenicia and Judaea, namely Bambyce, also called in Aramaic/Syriac
Manbog, and soon to be called in Greek Hierapolis, some miles west of the
Euphrates. This is the place from which came the slave in Egypt with his wrist
tattooed in ‘‘barbarian letters’’ (text to n.  above). It may be worth putting
together what we know of this place, somewhat increased since Goossens’
book of . The location of Bambyce, not far from the Euphrates and
northern Mesopotamia, may well be significant. Since we know even less of
the culture of northern Mesopotamia in the Hellenistic period than we do
of the various areas of Syria itself, we can only speculate about how far the
. See esp. Rey-Coquais (n. ).
. See esp. T. Rajak, Josephus: The Historian and His Society ().
. For the dynasty, see R. D. Sullivan, ‘‘The Dynasty of Emesa,’’ ANRW II. (), –
; J. Wagner, ‘‘Dynastie und Herrscherkult in Kommagene: Forschungsgeschichte und
neuere Funde,’’ Istanbuler Mitteilungen  (): ff.
. Waldmann (n. ).
. Schürer, Vermes, and Millar, History I, ff.
The Problem of Hellenistic Syria

two areas shared a common cultural history. But it is at least clear that a remarkable variety of non-Greek influences steadily gained ascendancy, from
the first century .. onwards, in the Macedonian colony of Dura-Europos;
that Syriac script is first attested, in the very early first century .., on the
Euphrates;78 and that another Macedonian colony, not far across the Euphrates, Edessa, was to be the focus of Syriac culture.79 For the wider cultural
contacts around the Fertile Crescent, it is suggestive that Lucian’s account of
the ‘‘Syrian Goddess’’ of Hierapolis/Bambyce records (dea Syr. ) that offerings came there not only from various regions west of the Euphrates but also
from Babylonia.
At the time of Alexander’s conquest Bambyce was producing coins with
Aramaic legends, with the names of ‘‘ ‘Abdhadad’’ (meaning ‘‘servant of [the
god] Hadad’’), or ‘‘ ’Abyaty,’’ or (still in Aramaic letters) ‘‘ ’Alksandr [Alexandros].’’ One has a longer legend ‘‘ ‘Abdhadad, priest of Manbog, who(?) resembles Hadaran his lord.’’ 80 The reverse of this same coin shows a priest,
presumably ‘Abdhadad himself, standing before an altar wearing a long tunic
and tall conical hat. Other coins represent Atargatis of Bambyce, the ‘‘Syrian
goddess,’’ and one of these has, in Greek, the letters ΣΕ, presumably ‘‘Seleukos.’’ According to Lucian the temple which stood there in his time had been
rebuilt by Stratonice, the wife of Seleucus I (dea Syr. ).
In the next century, inscriptions of / .. onwards record men from
this place, described as a polis with the name Hierapolis (e.g., ID no. ),
acting as priests of Hadad and Atargatis at Delos, where a whole range of
Syrian cults are represented explicitly in a way which is hardly attested in
the Hellenistic period in Syria itself.81 Probably a little earlier, an inscription
from Larisa in Thessaly reveals a man called Antipater, a ‘‘Hierapolitan of Seleucis,’’ described as ‘‘a Chaldaean astronomer,’’ evidently resident over a long
period in Thessaly. The description of him as a ‘‘Chaldaean,’’ later repeated
by Vitruvius,82 would naturally suggest either that Hierapolitans were felt to
. J. Pirenne, ‘‘Aux origines de la graphie syriaque,’’ Syria  (): –; H. J. W.
Drijvers, Old Syriac (Edessean) Inscriptions ().
. J. B. Segal, Edessa: The Blessed City (); H. J. W. Drijvers, ‘‘Hatra, Palmyra, und
Edessa: Die Städte der syrisch-mesopotamischen Wüste in politischer, kulturgeschichtlicher und religionsgeschichtlicher Bedeutung,’’ ANRW II. (), –.
. In general, see H. Seyrig, ‘‘Le monnayage de Hiérapolis de Syrie à l’époque d’Alexandre,’’ RN  (): –, who recalls the image in Ecclesiasticus , of the high priest Simon
as he emerged before the people from within the Temple; but, as Seyrig recognises, the
reading and interpretation are not certain.
. Bruneau (n. ), ff.
. G. W. Bowersock, ‘‘Antipater Chaldaeus,’’ CQ  (): .

The Hellenistic World and Rome
be in some way associated with Babylonia, or that his astronomical learning
was acquired there, or both.
In the first century .. the wealth of the temple was evidently well
known, and in  .. Crassus removed treasures from it (Plutarch, Crassus
). Under the Roman Empire one of the few inscriptions from the site (IGLS
I nos. –) shows that the place had a boulē and dēmos (council and people)
in the normal way (no. ). But the most striking of all the evidence is the
relief, in two halves found fifty years apart and joining perfectly, showing a
priest in a long tunic and conical hat surmounted by a crescent.83 A Greek
inscription records that this is a statue of Alexander, ‘‘the incomparable high
priest,’’ put up by his friend Achacus, who offered libations and prayed to the
gods to preserve his patris (homeland) in eunomia (good order). The statue
dates to the second century .., the time when Lucian describes the cult of
the goddess and lists the vestments of the various priests—the others in white
robes and pointed hats, the high priest in purple robes and a tiara (which
is visible in the relief, round the bottom of his tall hat). In this case there
is enough evidence to show a non-Greek cult which was already in existence before the Hellenistic period, and continued in a closely similar form
into the Roman Empire. Very early on in the Hellenistic period it seems to
have gained royal patronage; in the next century its cult is on show in Delos; under the Roman Empire it is a curiosity and tourist attraction, and a
suitable subject for Lucian’s parody of Herodotus.84
Outside Phoenicia and Judaea there is nowhere else in Seleucid Syria of
which we can say the same. Those few non-Greek, or mixed Greek and nonGreek, cultures which our evidence at present does allow us to observe either
came from outside the area of Seleucid control or are creations of the very
late Hellenistic and the Roman period, or both. By contrast with the dearth
of Aramaic inscriptions of the Achaemenid, Hellenistic, and Roman periods
from Syria, inscriptions in the various pre-Islamic Semitic scripts are known
in large numbers and cover a considerable range in space and time.85 Firstly,
Thamudic inscriptions begin in north-west Arabia around  .. and continue until the third or fourth century .. The sub-category of them known
as ‘‘Safaitic,’’ named from the volcanic region called the Safa, south-east of
Damascus, was in use from the second century .. to just before the rise of
. R. A. Stucky, ‘‘Prêtres Syriens II: Hiérapolis,’’ Syria  (): –.
. See now, for all these questions, J. L. Lightfoot, Lucan, On the Syrian Goddess ().
. H. P. Roschinski, ‘‘Sprachen, Schriften und Inschriften in Nordwestarabien,’’ BJ 
(): ff.; J. Teixidor, ‘‘L’hellénisme et les ‘Barbares’: l’exemple syrien,’’ in Le temps de la
réflexion (), –.
The Problem of Hellenistic Syria

Islam; scattered examples have been found as far away as Dura-Europos and
Hama. Secondly, the Nabataeans, whom our classical sources regard as Arabs,
could already write in ‘‘Syrian letters’’ when Antigonus made his unsuccessful campaign against them in  .. (Diod. Sic. , , ). Nabataean inscriptions, of which some , are known, begin in the first half of the second
century .., and continue until the fourth century ..86 The northward
spread of the inscriptions mirrors the spread of Nabataean control, which for
a period in the first century .., and possibly again in the first century ..,
included Damascus. Thirdly, as regards the parallel case of the Palmyrenes,
there is evidence of continuous occupation from the third millennium onwards on the tell where the temple of Bēl stood, and a mud-brick temple
may have been constructed there in the early Hellenistic period. But a tomb
of the mid-second century .. seems to be the earliest datable Hellenistic
structure on the site.87 In the middle of the first century .. we find both
the earliest Palmyrene inscriptions and the earliest evidence of monumental building.88 Both of these cases, Nabataea and Palmyra, can be argued to
be examples of the sedentarisation of Arab, or at least nomadic, peoples, and
certainly involved the construction of new urban centres exhibiting highly
distinctive local varieties of Greek architecture. Fourthly, a settled population of mixed Greek and non-Greek culture, with buildings and inscriptions,
also emerges in the same period in the Hauran (Djebel Druze), south-east of
Damascus. The earliest known monument there is the temple of Balshamen
at Si’a, dated by a bilingual Greek/Nabataean inscription to / ..89
It would be absurd to pretend that we can in any way explain these closely
parallel developments. All I wish to underline is that we can see the visible
manifestations of a number of mixed cultures emerging first outside the areas
of Seleucid, or Roman, control, and then spreading inwards. Or so it seems;
at Baalbek/Heliopolis, a place which we would naturally think of as distinctively Syrian, there is no certain archaeological evidence from before the
early Roman imperial period. Von Gerkan did however argue that under the
major temple of the mid-first century .. there were the foundations of
. J. Starcky, ‘‘Pétra et la Nabatène,’’ SDB  (), –; G. W. Bowersock, Roman
Arabia ().
. H. Seyrig, ‘‘Antiquités syriennes, : les dieux armés et les Arabes en Syrie,’’ Syria
 (): –; M. A. R. Colledge, The Art of Palmyra (); R. Fellmann, Le sanctuaire de Balshamin à Palmyre V: Die Grabanlage (); E. Will, ‘‘Le développement urbain de
Palmyre: temoignages épigraphiques anciens et nouveaux,’’ Syria  (): –.
. Drijvers (n. ).
. J.-M. Dentzer and J. Dentzer, ‘‘Les fouilles de Si’ et la phase hellénistique en Syrie
du sud,’’ CRAI (): –.

The Hellenistic World and Rome
a late Hellenistic temple of different design.90 Emesa, further north up the
Orontes valley, was also of course, at least by the second and third centuries,
the site of the conspicuously non-Greek cult of Elagabal, whose cult object
was an aniconic black stone. The place is not known to have existed until
the first century .., when there appeared the local dynasty of Sampsigeramus and his son Iamblichus,91 the ‘‘tribal leaders’’ ( phylarchoi) of the people
(ethnos) of the Emiseni, as Strabo calls them (, ,  []), saying that they
ruled nearby Arethusa (a settlement of Seleucus I, Appian, Syr. ). These
dynasts too were characterised by contemporaries as ‘‘Arabs’’ (Cicero, fam. ,
, : ‘‘Iamblichus, phylarchus Arabum [tribal leader of the Arabs]’’), and the
first part of Sampsigeramus’ name is based on the Semitic word shemesh, the
sun. On Seyrig’s view, the sun cult in Syria is typically of Arab origin.92 But,
just to confuse our conception of the background, the main cult of Emesa
in this period does not itself seem to have been a sun cult. What seems to
be the earliest documentary attestation of the name ‘‘Elagabal’’ as a divine
name offers a new etymology for the word, namely ‘‘god mountain.’’ This is
an inscription in Palmyrene lettering of the first century .., found some 
kilometres south-east of Emesa and  kilometres south-west of Palmyra,
and naming, along with the Arab deity Arṣu, another deity called ‘‘ ’Ilh’ gbl,’’
that is (perhaps), ‘‘Elaha Gabal’’—‘‘god mountain’’—represented as an eagle
with outstretched wings standing on a rock.93 Then, to add a further confusing element, when we come to Herodian’s famous description of the cult
of Elagabal (, , –), he characterises it as ‘‘Phoenician’’; just as Heliodorus,
the author of the novel Aethiopica, calls himself ‘‘a Phoenician from Emesa’’
(, , ).
However we ought to characterise the cultural background out of which
the Emesa of Roman imperial times emerged, Seyrig elsewhere saw its brief
prosperity as a city as having been closely linked to the caravan trade of Palmyra.94 That raises questions which cannot be dealt with here. All I wish to
emphasise is that there is nothing to show that Emesa or its cult even existed
. A. von Gerkan, ‘‘Die Entwicklung des grossen Tempels von Baalbek,’’ Corolla Ludwig
Curtius zum . Geburtstag dargebracht () ff. E. Boehringer, ed., Von antiker Architektur
und Topographie: Gesammelte Aufsätze (Stuttgart, ), –.
. Sullivan (n. ).
. H. Seyrig, ‘‘Antiquités syriennes, : le culte du Soleil en Syrie à l’époque romaine,’’
Syria  (): –.
. J. Starcky, ‘‘Stèle d’Elahagabal,’’ Mélanges Université St.-Joseph  (–): ff.
. H. Seyrig, ‘‘Antiquités syriennes, : caractères de l’histoire d’Emése,’’ Syria  ():
–.
The Problem of Hellenistic Syria

in the Hellenistic period proper. One hypothesis is to see its emergence as
a product of the movement of ‘‘Arabs’’ inwards from the desert fringes, followed by their settlement and creation of a new cult, or at any rate one which
was new to that site.
If we move somewhat further north, to Seleucus’ foundation at Apamea,
here in the Roman Empire there was a cult of a non-Greek deity, whom Dio
(, , ) describes as ‘‘Jupiter called Belos, who is worshipped in Apamea
in Syria’’ and who gave oracular responses. The god Bēl, worshipped also in
Palmyra,95 is first attested in Babylonia. How and when the cult had come
to be set up in a Greek city, or to be associated with a cult already there,
we do not know.96 But in the entire range of our evidence there is probably
no more concentrated example of cultural fusion than the brief inscription
from Apamea published by Rey-Coquais,97 a Greek dedication by a Roman
citizen: ‘‘On the order of the greatest holy god Bēl, Aurelius Belius Philippus,
priest and diadochos [successor] of the Epicureans in Apamea.’’
The enigma of Hellenistic Syria—of the wider Syrian region in the Hellenistic period—remains. None the less, I am tempted to speculate that the positive impact of Hellenistic rule was relatively slight. If we think of it in terms
of the foundation of wholly new cities, these were not numerous, except in
northern Syria, and only a few of them are known to have closely resembled
what we think of as a fully fledged Greek city. If we think of an economic
or social impact, there were many areas where the Seleucid Empire certainly
never exercised any direct or effective control.98 What the Seleucid state did
was to raise taxes where it could, and to enrol troops either (perhaps) by direct levies among Macedonian klērouchoi (colonists) or, more probably, via
the Greek cities, like Cyrrhus, via local dynasts like the Hasmoneans, who
from time to time supplied contingents, or were supposed to ( Macc. :;
:), or from Arab dynasts like Zabdibelus, who led , Arabs at the
battle of Raphia in  (Polybius , , ). The Seleucid state, like most ancient states, was primarily a system for extracting taxes and forming armies.
Much of Syria was disputed territory between the Seleucid and Ptolemaic
. J. Teixidor, The Pagan God: Popular Religion in the Ancient Near East (), –.
. Balty and Balty (n. ), , n. C; J. Balty, ‘‘L’oracle d’Apamée,’’ AC  (): –.
. J.-P. Rey-Coquais, ‘‘Inscriptions grecques d’Apamée,’’ AAAS  (): .
. But see the preface by the editors to Kuhrt and Sherwin-White (n. ), ix–xii; S. M.
Sherwin-White, ‘‘Seleucid Babylonia: A Case Study for the Installation and Development
of Greek Rule,’’ in Kuhrt and Sherwin-White (n. ), –; J.-F. Salles, ‘‘The Arab-Persian
Gulf under the Seleucids,’’ in Kuhrt and Sherwin-White (n. ), –.

The Hellenistic World and Rome
kingdoms throughout the third century .. Antiochus IV’s final invasion of
Egypt in  had as an immediate consequence the desecration of the Temple
in Jerusalem (or so it seemed to one contemporary, Daniel :–), and the
imposition of a Seleucid garrison. A mere six years later, with the escape of
Demetrius I from Rome, there began a series of civil wars over the succession to the Seleucid throne which did not end until the occupation of Syria
first by Tigranes of Armenia and then by the Romans.
The nature of the Seleucid state, as seen by its subjects, is suggested by the
importance of the right of asylia (asylum) as granted to cities,99 just as it is
by Ptolemaeus’ concern, immediately after the conquest of southern Syria,
to have his villages protected from billetting by the Seleucid army (see text
following n.  above). Some decades later, after the death of Antiochus VII
Sidetes while on campaign against the Parthians in , his stratēgos (general) Athenaeus, when in flight, was refused entry or supplies by the villages
which had been ‘‘wronged in connection with the epistathmeiai [quartering
of soldiers]’’ (Diod Sic. /, , ).
It is worth suggesting the hypothesis that the remarkable absence of tangible evidence from Syria in the Hellenistic period may not be an accident
which further discovery would correct, but the reflection of a real absence
of development and building activity in an area dominated by war and political instability. Given this absence of evidence, we cannot expect to know
much about the culture of Syria in this period, or whether there was, except
along the coast, any significant evolution towards the mixed culture which
came to be so vividly expressed in the Roman period. The hints which we
gain of such a culture are hardly worth mentioning: for instance the fact
that Meleager of Gadara, whose epigrams are entirely Greek in spirit, at least
knew what words were used as expressions of greeting both in Aramaic and
in Phoenician.100 But there is nothing in the quite extensive corpus of his
poetry to show that he had deeply absorbed any non-Greek culture in his
native city, although no formal Greek or Macedonian settlement is attested
there.101 On the contrary, he self-consciously represented his native city as
‘‘Attic Gadara situated among the Assyrioi,’’ and says of himself ‘‘If [I am] a
Syrian, what is the wonder? My friend, we inhabit a single homeland, the
world.’’ 102 For evidence of non-Greek culture on the part of the inhabitants
. E. Bickermann, Institutions des Séleucides (), .
. Anth. Pal. .; A. S. F. Gow and D. L. Page, The Greek Anthology: Hellenistic Epigrams
(), , no. iv.
. Schürer, Vermes, and Millar, History II, –.
. Anth. Pal. .; Gow and Page (n. ), , no. ii.
The Problem of Hellenistic Syria

of inland Syrian cities in the Hellenistic period, one can add a passing allusion to the fact that Antonius could find in Antioch in the s .. a leading
citizen who knew Aramaic, or perhaps Parthian (Plutarch Ant. ).
One of the major problems in the understanding of Hellenistic Syria is
thus the relative scarcity of direct and contemporary evidence for any nonGreek culture, or cultures, in the region, either in the Achaemenid or the
Hellenistic period itself. That might not matter, if we were confident that
the evidence available for the Roman imperial period could be used to show
cultural continuity, rather than the importation of new elements, from the
desert, from Babylonia, or from Mesopotamia. The question of chronology
may be crucial, and certainly cannot be ignored. To give one central example, in his famous book of , Der Gott der Makkabäer, Bickerman argues
that we should envisage the pagan cult imposed in  .. on the Temple in
Jerusalem not as Greek but as Syrian. In particular he explains the emphasis
which Jewish sources place specifically on the desecration of the altar by the
‘‘abomination of desolation,’’ by the parallel of Arab worship of the altar as a
cult object in itself. To reinforce this, he uses the example of an inscription
from Jebel Sheikh Barakat near Beroea (Aleppo) with a dedication to Zeus
Madbachos, ‘‘Zeus of the Altar.’’ But there is an acute problem of chronology
here: the temple from which this inscription comes did not exist in the second century .. It was constructed by persons with Greek names between
the s and the s ..; the earliest inscriptions recording its dedication to
Zeus Madbachos and Selamanes, ‘‘the ancestral gods,’’ probably date to the
s ..103
The ancestors of these people may indeed have worshipped these same
deities through the Hellenistic period. The god Shulman/Selamanes is in fact
attested in Syria long before that. But nobody, so far as we know, put up a
temple for these gods on this site, or composed a dedicatory inscription for
them until the first century .. The problem therefore remains. Whatever
the society, economy, and culture of the Syrian region was like in the Hellenistic period, the ‘‘Hellenistic’’ Syria, with a distinctive mixed culture, which
our evidence allows us to encounter is that which evolved under the Roman
Empire.104
That is, however, in the first instance, a fact about our evidence. It is not
presented here as a definite conclusion about the ‘‘real’’ world of the Syrian
. IGLS II, nos. –; see O. Callot and J. Marcillet-Jaubert, ‘‘Hauts-lieux de la Syrie
du Nord,’’ Temples et Sanctuaires (), ff.
. See esp. Teixidor (n. ), for the popular religion attested in the inscriptions of this
period.

The Hellenistic World and Rome
region in the Hellenistic period, but as a strategic device whose purpose is
precisely to bring into sharper relief significant new items of evidence as
they appear. Firstly, to insist on the sparseness of evidence for the culture and
social structure of the region in the Achaemenid period, with the possible
exception of Judaea, and to a lesser extent Phoenicia, is to prevent the unconscious projection of general notions about ‘‘oriental’’ or ‘‘Near Eastern’’
civilisation on to this area. Secondly, to emphasise the limits of the empirical
data which we can actually use to give substance to the notion of ‘‘Hellenisation’’ or ‘‘Hellenism’’ in this particular time and place is both to call these
concepts into question and to insist on testing them, so far as possible, area
by area and period by period. Thirdly, the notion of a ‘‘fusion’’ of cultures is
doubly open to question if we have very little direct evidence for the nature
of either of the cultures concerned, let alone for the manner in which they
may have interacted, or occupied separate spheres. ‘‘Hellenisation’’ might, as
is often supposed, have extended very little outside the towns or the upper
classes. Yet as regards towns, or urban centres, there is enough evidence to
suggest that it was possible to absorb Greek culture without losing local traditions; and that Hierapolitans, Phoenicians, and Samaritans when abroad
positively emphasised their non-Greek identity.
Nor, by contrast, is it certain that country areas remote from the centres
of Greek or Macedonian settlement remained immune to Greek presence or
influence. This paper concludes with what seems to be (so far) the only formal
bilingual inscription, in Greek and Aramaic, dating to the Hellenistic period,
and discovered west of the Euphrates. This is a dedication from Tel Dan, first
published in a brief archaeological report and discussed by Horsley in one
of his valuable surveys of new material relevant to early Christianity.105 The
site seems to have been a high place of the Israelite period (tenth–ninth centuries ..), on which further construction, possibly including an altar, subsequently took place in the Hellenistic period. The inscription, carved on a
limestone slab, seems to date to the late third or early second century ..
(BE , no. ). The Greek text, quite finely carved, presents no problems:
‘‘To the god who is in Dan Zoilos (offers) his vow’’ (theōi / tōi en Danois /
Zōilos euchēn). Immediately underneath it comes an Aramaic text, more amateurishly carved, of which just enough survives to show that the author, and
hence the date, is the same. It reads either [BD]N NDR ZYLS L’[LH’]—‘‘In
. A. Biran, ‘‘Chronique archéologique: Tell Dan,’’ RB  (): –; G. H. R.
Horsley, New Documents Illustrating Early Christianity I: A Review of the Greek Inscriptions and
Papyri Published in  (), no. .
The Problem of Hellenistic Syria

[Da]n, vows of ZYLS (Zoilos) to the god,’’ or (more probably) [H]N NDR
ZYLS L’[LH’ DN]—‘‘[This] (is the) vow (of ) Zoilos to the [god in Dan].’’
On either interpretation, this modest document is of immense significance for the cultural and religious history of the Syrian region. Firstly, it is
one of the earliest formal Greek inscriptions from the whole area. Secondly,
it is both the only formal Aramaic (as opposed to Phoenician) inscription and
the only formal Greek-Semitic bilingual inscription (as opposed to ostraca)
from the Syrian region in the Hellenistic age. Thirdly, the archaeological evidence clearly suggests the continuation, or at least the resumption, of worship
at an ancient cult site. Fourthly, the site itself occupies an inland location,
near the headwaters of the Jordan, separated from the coast by some forty
kilometres of hill country, and some fifty kilometres away from the nearest Greek, or semi-Greek, cities, Damascus and Gadara. There is no way of
knowing whether Zoilos was an immigrant Greek who had either acquired
some knowledge of Aramaic, or at least knew the necessity of having his vow
recorded also in Aramaic; or whether he was a person of Syrian origin who
had learned Greek, and adopted the Greek custom of the dedicatory inscription, and paired it with an inscription of a less well established type, in his
native Aramaic. In either case a rustic cult centre saw worship directed to its
nameless deity and recorded in a fine Greek inscription. Here at last we have
a precise example, from the earlier Hellenistic period, of the meeting of two
identifiable cultures.
 
The Phoenician Cities:
A Case-Study of Hellenisation *
When Alexander was civilising Asia, Homer was commonly read, and
the children of the Persians, of the Susianians and of the Gedrosians
learned to chant the tragedies of Sophocles and Euripides. And although Socrates, when tried on a charge of introducing foreign deities,
lost his cause to the informers who infested Athens, yet through Alexander Bactria and the Caucasus learned to revere the gods of the Greeks.
Plato wrote a work on the one ideal constitution, but because of its
forbidding character he could not persuade anyone to adopt it; but
Alexander established more than seventy cities among savage tribes and
sowed all Asia with Grecian magistracies, and thus overcame its uncivilised and brutish manner of living.
These familiar words of Plutarch (Mor.  D–E, Loeb trans.) begin to seem
not quite as foolish as they did, in the light of modern discoveries in Aï
Khanum and Kandahar. They may thus serve to raise some larger questions.
Firstly, it is curious how Plutarch concentrates on remote central Asian areas
which were no longer Hellenised in any obvious sense in his own day. Secondly, he emphasises, as we would expect, the creation of new cities with
Greek constitutions. Here we might well turn to a neglected passage of his
older contemporary, Josephus, concluding his account of the tower of Babel.
* First published in Proceedings of the Cambridge Philological Society  (): –. Earlier
versions were read at the Cambridge Philological Society in April  and subsequently
at the Institut für Alte Geschichte und Epigraphik in Munich and at the Istituto di Storia
Antica at Pavia. Since it cannot pretend in any case to be more than a sketch, it has been
left in the same form, with added annotation.

The Phoenician Cities

Of the peoples, some still preserve the names which were given to them
by their founders, some have changed them, while others have adopted
a form of name designed to be more intelligible to those who are settled
among them. It is the Greeks who are responsible for this. For when
they subsequently rose to power they appropriated to themselves even
the glory of the past, adorning the peoples with names which were intelligible to themselves and imposing on them a form of constitution
as if they were descended from themselves. (Ant. , , Loeb trans.)
A little later he gives a specific example: ‘‘Amathus . . . which is still called
‘Amathe’ by the local people [epichorioi], though the Macedonians named it
‘Epiphaneia’ after one of the epigonoi [Alexander’s successors]’’ (Ant. , ).
Jerome confirms that ‘‘until our own time’’ the town was still called Hamath
‘‘both by the Syrians [Syri] and by the Hebrews [Hebraei]’’ (Hebr. Qu. Gen.
, –, in CCL LXXII, ). It is of course an unquestionable fact that
there was large-scale city foundation by the Seleucids, primarily Seleucus I,
in North Syria, in the Orontes valley, and in Mesopotamia. I need only refer
to Seyrig’s classic study of the urbanisation of North Syria by Seleucus I.1
The effect in both areas was to produce, in the first instance, societies with
two separate cultures: the Aramaic culture of Hellenistic Syria remains, it
is true, almost invisible to us; the Akkadian culture of Babylonia, using the
cuneiform script, was to survive at least until the first century .. It is not irrelevant to the present study that it produced one native-born interpreter in
Greek, Berosus of Babylon, in the early Hellenistic period.2 It was this world
of contrasting cultures in Mesopotamia and Syria, cultures whose nature and
functioning in the Graeco-Roman period remain almost wholly unintelligible, which eventually produced something like a fusion of cultures, in the
form of Syriac Christianity. To illustrate this point I need only mention that
Syriac literature originally stemmed from a Macedonian colony, Edessa.
All recent studies of Hellenisation or ‘‘Hellenism’’ have emphasised that
a fusion of Greek and native cultures was categorically not what Greeks of
the fourth century and after had intended. We have all learnt from Momigliano’s Alien Wisdom () how slight an interest Greeks took in other cultures. Martin Hengel in Jews, Greeks and Barbarians () and Claire Préaux
. H. Seyrig, ‘‘Séleucus I et la fondation de la monarchie syrienne,’’ Syria  (): .
Note also the important study, by P. Briant, ‘‘Colonisation hellénistique et populations indigènes: la phase d’installation,’’ Klio  ():  P. Briant, Rois, tributs et paysans: études
sur les formations tributaires du Moyen-Orient ancien ().
. FGrH . See S. M. Burstein, The Babyloniaca of Berossus (Sources and Monographs,
Sources from the Ancient Near East, ., ).

The Hellenistic World and Rome
in her great masterpiece, Le monde hellénistique I–II (), as previously in
‘‘Réflexions sur l’entité hellénistique,’’ Chron. d’Eg.  (): , have made
clear that the conscious expressions of Greek opinion available to us envisage either an export of Greeks and their culture or at the most an adoption
of that culture by ‘‘barbarians’’; and that is of course also the retrospective
view of Plutarch, with which I began. Pierre Briant (n. ) has also argued that
the Hellenistic foundations did in fact function, at least in the initial phase,
as nuclei of social segregation and dominance in relation to the indigenous
populations.
These analyses in effect seek—more or less successfully—to remove Droysen’s notion of Verschmelzung from the logical field of the intentions of the
actors in the fourth century and Hellenistic period. They are of considerable importance, even if—as is obvious—we cannot and do not know how
the main actors, Alexander and the Seleucid kings, conceived the role of
the cities which they founded. But two points arise in regard to Phoenicia in particular. Firstly, since the foundation of Greek cities is central to
the process of Hellenisation ‘‘from above,’’ it is important to emphasise that
such foundations, if they took place at all, were not characteristic of coastal
Phoenicia. Of the cities along the coast which Pseudo-Scylax listed in the
mid-fourth century (GGM I, –), almost all preserved their ancient nonGreek names and were not refounded—Arados, Tyre, Sidon, Botrys, Berytus,
Sarepta, Ake, Doros, Joppa (so also Byblos, which this confused text does not
mention). There is one clear marginal exception: Callimachus, coins, and the
Zenon papyri all show that Ake had become ‘‘Ptolemais’’ by the s ..;3
there is also another possible (and temporary) exception, if coins of the second century .. with the legend in Phoenician ‘‘Laodicea, which is in [or metropolis of ] Canaan’’ are rightly attributed to Berytus.4 It is also a curious fact
that Tripolis already had this Greek name when Pseudo-Scylax was writing;
what is more he explains it by the fact that the city was in some way tripartite,
divided among Arados, Tyre, and Sidon. None the less, it has been suggested
that a Phoenician root ṬRP with the meaning ‘‘grasp,’’ or here ‘‘new foundation,’’ might lie behind this term. The explanation is speculative, to say the
least, but cannot wholly be ruled out.5
. For the evidence Schürer, Vermes, and Millar, History II, .
. BMC Phoenicia ff., –. See R. Mouterde, ‘‘Regards sur Beyrouth phénicienne, hellénistique et romaine,’’ Mél. Univ. St. Joseph  (): .
. K. Galling, ‘‘Die syrisch-palästinische Küste nach der Beschreibung des PseudoSkylax,’’ Studien zur Geschichte Israels im persischen Zeitalter (), . The question of the
name is not discussed in the valuable article by J. Elayi, ‘‘Studies in Phoenician Geography
during the Persian Period,’’ JNES  (): –.
The Phoenician Cities

Phoenicia thus provides a very clear exception to Josephus’ general picture. No Greeks colonised Arados, Tyre, Sidon, or Byblos or re-named them
in ways intelligible to themselves. Did any Seleucid or Ptolemaic king give
any of them a new, Greek constitution? That is a question which we must
postpone for a moment.
At all events the traces of an organised, deliberate effort of Hellenisation in Phoenicia are markedly slight. But even if no Greeks intended to
‘‘Hellenise’’ Phoenicia, such a Hellenisation might still have occurred. What
is more, from the fact that (as it seems) no Greeks sought a Verschmelzung
in Droysen’s sense, it does not in the least follow that no such fusion took
place. If we take only the two most obvious criteria, language and architectural forms, then we have two perfectly clear cases of mixed cultures which
evolved in the Hellenistic period on the desert fringes of the Syrian region:
in Nabataea from the early Hellenistic period onwards and in Palmyra from
the late Hellenistic period onwards.6
The case of Phoenicia, however, has two extra dimensions which make it
quite different from any other area in the Near or Middle East. The first, and
more important, is the surviving effects of Phoenician colonisation. If I may
digress for a moment, our view of ancient Mediterranean history remains, in
spite of all our efforts, firmly Graeco-Roman. But we really could now write
a Phoenician-Punic orientated history of the Mediterranean from the eighth
century onwards, which would focus on the nature, spread, survival, and decline of Phoenician-Punic culture in Phoenicia itself, in Cyprus, in North
Africa, in Sicily, in Spain, in Sardinia, and marginally in Italy.7 We are not
talking about an insignificant or short-lived phenomenon. Let us take for instance a neo-Punic inscription from Bitia in Sardinia, which deserves a special place among the quite extensive Punic and neo-Punic inscriptions from
the western Mediterranean. This is a building inscription, dated by a ruler
whose name, written in neo-Punic letters, comes out as [’MP]R’ṬR Q’YSR
M’RQH ’ WRHLY ’NṬNYNH [’]WGSṬH—‘‘Imperator Caesar Marcus Aurelius Antoninus Augustus,’’ namely Marcus Aurelius, Caracalla or Elaga. For Nabataea, see Schürer, Vermes, and Millar, History I, app. ; Y. Meshorer, Nabataean Coins (Qedem III, ); M. Lindner, ed., Petra und das Königreich der Nabatäer 2 ().
For Palmyra, see M. A. R. Colledge, The Art of Palmyra ().
. For Phoenician settlement, note C. R. Whittaker, ‘‘The Western Phoenicians: Colonisation and Assimilation,’’ PCPS  (): , and the important volume edited by H. G.
Niemeyer, Phönizier im Westen (Madrider Beiträge VIII, ). For a sketch of the role of
Punic in the western Mediterranean in the Roman period, see W. Röllig, ‘‘Das Punische im
römischen Reich,’’ in G. Neumann and J. Untermann, eds., Die Sprachen im römischen Reich
der Kaiserzeit (), .

The Hellenistic World and Rome
bal.8 It does not matter which. For what the inscription shows is that at least
four centuries after the Roman acquisition of Sardinia, Punic was still known,
was used, and could be composed and written. We perhaps talk too glibly of
the Romanisation of the western Mediterranean. It surely depended more
than we have allowed on a previous Hellenisation or Punicisation (or whatever the word should be). Punic culture survived not only in Sardinia; in
Africa Punic could be written until the second or third centuries, in neoPunic, Latin, or Greek characters, and was spoken, at any rate as a peasant
language, at least until the end of the fourth century ..9 When Augustine’s friend Valerius heard some peasants near Hippo use the word ‘‘shalosh’’ to mean three, he was very interested—but only because the concept
‘‘three’’ meant the Trinity and ‘‘shalosh’’ sounded hopefully like salus, salvation (PL XXXV, ). For our purposes it is a pity that he was not more
intrigued by the fact that Punic had remained in use for no less than five
and a half centuries after the end of Carthage—and that the peasants, when
questioned, duly identified themselves as Chanani, ‘‘Canaanites.’’
I use these examples mainly as models, to raise the question of what might
have happened in Phoenicia proper. But in the Hellenistic period at least
there is also some direct connection with the western colonies, a connection which was maintained unbroken through the period of Macedonian
conquest. Everybody knows that the Phoenicians firmly declined Cambyses’
order to sail against Carthage (Herodotus , ); but we may also recall that
when besieged by Alexander the Tyrians sent at least some non-combatants
to Carthage (Diodorus , , –). Moreover, when Alexander took the
place, he captured there some Carthaginian envoys who had come ‘‘in accordance with an ancient law’’ to worship Herakles/Melqart (Arrian, Anab. ,
, .). Publication is also awaited of a fourth-century .. Punic funeral stele
from near Tyre, of a native of Carthage who evidently died there.10 But far
more important is Polybius’ splendid story of the escape of the Seleucid king
Demetrius I from Rome in , in which he himself assisted; for the ship on
which Demetrius travelled was a Carthaginian one on its way to take firstfruits to Tyre (, , –). If this obligation was maintained a mere sixteen
. For the text, see M. G. Guzzo Amadasi, Le iscrizioni fenicie e puniche delle colonie in
Occidente (), , no. ; KAI, no. .
. See F. Millar, ‘‘Local Cultures in the Roman Empire: Libyan, Punic and Latin in Roman Africa,’’ JRS  ():  ( chapter  of Fergus Millar, Rome, the Greek World, and
the East II: Government, Society, and Culture in the Roman Empire); see also Röllig (n. ).
. This is reported by J. Teixidor in ‘‘Bulletin d’épigraphie sémitique,’’ , no. ,
published in Syria .
The Phoenician Cities

years before the fall of Carthage, we can safely assert that the links between
mother-city and colony never were broken.
These links have in any case to be seen against the background of the
continued presence of Tyrians in the western Mediterranean; for instance,
in the third or second century .. two Tyrian brothers made a dedication
in Malta to Herakles/Melqart, the inscription being bilingual in Phoenician
and Greek.11 Nor was it ever forgotten that the Herakles whom Hellenised
Tyrians worshipped (e.g., Inscriptions de Délos, no. , from / ..) was
really Melqart; this identification, already present in the pages of Herodotus,
is clearly asserted by Lucian, dea Syra . It was also surely not unknown to
the Tyrian who in Lucian’s time, to be precise .. /, made a dedication
to theos hagios (the holy god) Herakles; a man by the name of Diodorus, son
of Nithumbalus. It can hardly be an accident that the Greek name of the son
has precisely the same meaning—‘‘gift of god’’—as the Phoenician name of
the father.12
That brings me to the second dimension. Phoenician culture seems in
some sense to have spread inland as well as overseas in the Hellenistic period,
as Punic culture did also in North Africa after  .. This fact brings Phoenician culture into connection with the familiar phenomenon of the fusion
of Greek and non-Greek deities in Syria, or alternatively the survival of nonGreek cults in a Hellenised environment. There is nowhere where it appears
more vividly before us than in Herodian’s description of the cult of Elagabal
at Emesa (, , –); what is significant is that Herodian thought that ‘‘Elagabal’’ was a Phoenician name and that Julia Maesa was ‘‘by origin a Phoinissa’’;
it is relevant that Heliodorus, the author of the Aethiopica, also describes himself (, , ) as a ‘‘Phoenician from Emesa.’’ Similarly, according to Josephus, in the s .. the Samaritans identified themselves as ‘‘Sidonians in
Shechem’’ and asked that the nameless God to whom their temple was dedicated should be called ‘‘Zeus Hellenios’’ (Ant. , –). On the one hand,
this evidence provides the essential analogy on which Bickerman based the
view put forward in Der Gott den Makkabäer: that what the Maccabees were
reacting against was a conscious reform movement within Judaism, designed
to preserve the cult of the Highest God in a form acceptable to the rest of the
Hellenistic world. On the other hand, it is further evidence, like the wellknown inscriptions (in Greek) of the ‘‘Sidonians in Marissa’’ in Idumaea,13
. Guzzo Amadasi (n. ), no. .
. See M. Chéhab, ‘‘Tyr à l’époque romaine,’’ Mél. Univ. St. Joseph  (): , on p. .
. OGIS . Note the re-examination of the extent of Hellenisation in Marisa by
G. Horowitz, ‘‘Town Planning of Hellenistic Marisa,’’ PEQ – (–): .

The Hellenistic World and Rome
that people in the inland areas could identify themselves as Phoenicians or
Sidonians. The history of the Jews, from the kingdom onwards, is in any case
to be seen as that of a small community in the hinterland, in the hills of
Judaea or Galilee, behind the Phoenician, and a bit later Phoenician-Greek,
cities of the coast; and the inland extension of Phoenician culture can only
have intensified contacts. Phoenician is also the closest Semitic language to
Hebrew; and to the very end Tyrian shekels were the standard currency in
which the Temple dues were to be paid.14 Jesus and his followers did not have
to travel very far to cross into the territory of Tyre and Sidon, where they
met a woman whom Mark (:) calls a ‘‘Greek woman, a Syro-Phoenician
by birth,’’ and Matthew (:) a ‘‘Canaanite.’’ By comparison with the hopeless confusingness of most ethnic labels from antiquity this contrasted pair
is actually quite informative.
Opinions of course differ on the question of how fundamentally the Judaism which Jesus represented had itself been Hellenised. I would still hold to
the view that what is significant is not such Hellenisation as there was, but
the maintenance of a historical, religious, and cultural tradition enshrined
in sacred books in Hebrew (which were however available, even in Judaea,
in Greek translation) and which could generate further religious works in
both Hebrew and Aramaic.15 Given the very close geographical, historical,
and linguistic links between Judaea and Phoenicia, we may reasonably allow
the vitality of Jewish historical and religious culture to pose the question as
to whether, or to what extent, the same might have been true of Phoenicia.
It is indeed at this level alone, or so it seems to me, that we can actually approach the question of the Hellenisation of the Phoenician cities: that is the
level of conscious historical tradition and identity. The archaeological record
for this period is extremely poor; nothing that I have found, at any rate,
has given me any clear conception of a material culture of Phoenicia in the
Persian period,16 which might or might not have been transformed by Hellenistic influences. Nor, I think, is it possible to discern anything about any
possible changes in the relations of production. Whether or not it is useful to
talk of an ‘‘Asiatic mode of production,’’ 17 the question is often asked whether
. See A. Ben-David, Jerusalem und Tyros: ein Beitrag zu palästinensischen Münz-und Wirtschaftsgeschichte,  a.c.– p.c. ().
. F. Millar, ‘‘The Background to the Maccabean Revolution: Reflections on Martin
Hengel’s Judaism and Hellenism,’’ JJS  ():  ( chapter  in the present volume).
. Compare, however, E. Stern, Material Culture of the Lands of the Bible in the Persian
Period, – .. ().
. See most recently L. Zaccagnini, ‘‘Modo di produzione asiatico e vicino Oriente
antico. Appunti per una discussione,’’ Dial. di Arch.  (): .
The Phoenician Cities

Hellenisation brought with it a system of production based on the ownership of slaves by private persons; or whether on the contrary no such change
took place, large-scale landownership, with the land being worked by dependent labourers, remained the rule, and therefore the Greek polis in Asia
bore no real resemblance to the original, as Kreissig has argued.18 Such fundamental questions of social structure seem difficult to relate to Phoenicia if
only because, from what little we know, the Phoenician cities seem already
to have exhibited a market economy based on private wealth, in which trade
was at any rate a distinctive feature. That is to say, they exhibited what has
been called Vorhellenismus, namely being rather like Greek poleis anyway.19
Nor have we enough evidence even to illustrate the effects of entry into
the Hellenistic world on ordinary life in Phoenicia. By contrast we have Aramaic ostraca of the third century .. from Idumaea, from Jerusalem, and
from Elath which already show extensive penetration of Greek loan-words
in economic life, even at this early date.20 In Phoenician, so far as I know, we
have no such non-formal written documents from this period. The nearest
we have are mere scraps of evidence: an inscribed jar from near Jaffa, probably of the later fourth century;21 or a badly damaged Phoenician papyrus
of unknown origin, now in the Cairo Museum, and thought to be of the
fourth or third century, which is in some way concerned with the delivery
of natural products.22
There seems at the moment to be no way of reaching any idea of changes
in the structure of ordinary life from the Persian to the Hellenistic period in
Phoenicia. What can be done is perhaps three things: to look at the continued
use of Phoenician as represented on formal inscriptions put up by individual
Phoenicians abroad in the Hellenistic period; to look at the evolution of the
cities as communities; and to ask if there was anything which resembled the
continuous historical and religious tradition which is so strikingly charac. H. Kreissig, ‘‘Die Polis in Griechenland und im Orient in der hellenistischen
Epoche,’’ in E. C. Welskopf, ed., Hellenische Poleis II (), .
. J. P. Weinberg, ‘‘Bemerkungen zum Problem ‘Der Vorhellenismus im Vorderen Orient,’ ’’ Klio  (): . For a very useful survey of the available evidence, note J. Elayi, ‘‘The
Phoenician Cities in the Persian Period,’’ Journal of the Ancient Near Eastern Society of Columbia
University  (): .
. Idumaea: L. T. Geraty, ‘‘The Khirbet el-Kôm Bilingual Ostracon,’’ BASOR 
(): . Jerusalem: F. M. Cross, ‘‘An Aramaic Ostracon of the Third Century ... from
Excavations in Jerusalem,’’ Eretz Israel  (): *. Elath: N. Glueck, ‘‘Ostraca from Elath,’’
BASOR  (): .
. B. Peckham, ‘‘An Inscribed Jar from Bat-Yam,’’ IEJ  (): .
. KAI, no. .

The Hellenistic World and Rome
teristic of Judaea. In other words if our evidence will tell us anything, it is
only at the level of conscious, formal expression.
If we were to set out laboriously to collect the epigraphic references to,
or by, citizens of the Phoenician cities abroad in the Hellenistic world, we
know anyway what we would find: that the majority appear as Greek in language and nomenclature.23 Nor would we learn anything significant if we
collected all the examples of Semitic names in use among these people, as
they certainly were throughout the classical period. It does, however, have
some significance, if only illustrative, that at least some Phoenicians abroad
continued to compose and inscribe texts in Phoenician. At Demetrias, for
example, three men from the third century .., all with Greek names—
one from Sidon, one from Arados, and one from Kition—have left brief inscriptions in Phoenician.24 In Athens we have a bilingual Greek-Phoenician
inscription giving a fine example of equivalence, or semi-equivalence, in
theophoric names. Artemidorus son of Heliodorus, a Sidonian, appears in his
Phoenician form as ‘BD TNT BN ‘BDŠMŠ. Tanit therefore equals Artemis,
as Shemesh (the Sun) certainly does Helios. But the root ‘BD means ‘‘slave,’’
not ‘‘gift’’ (as ‘‘-dorus’’ in ‘‘Heliodorus’’ does).25 On another bilingual inscription, probably of the third century, from the Peiraeus, the female name ’SPT
daughter of ’ŠMNŠLM is transliterated Ἀσεπτ Ἑσυμσεληου and a Phoenician dedication of an altar, perhaps of the second or first century, mentions
a man who was son of B‘LYṬWN, the ŠPṬ (shofet, i.e., ‘‘judge’’; see text to
n.  below).26 There is also a substantial eight-line Phoenician inscription,
accompanied by a one-line Greek text, from the Peiraeus, probably of the
mid-third century, and dated by ‘‘the th year of the people of Sidon,’’ which
shows the koinon (community) of the Sidonians there crowning a religious
official. Once again his Greek name, ‘‘Diopeithes,’’ is vaguely equivalent to
his Phoenician one, ŠMB‘L—‘‘God has heard.’’ 27 There are also, for instance,
three brief Phoenician-Greek bilingual inscriptions from Rhodes, probably
of the early third century.28
Taken together, this scatter of inscriptions is of some significance. Citi. For the case of Arados, see the collection by J.-P. Rey-Coquais, IGLS VII, –.
. O. Masson, ‘‘Recherches sur le Phéniciens dans le monde hellénistique,’’ BCH 
(): .
. KAI, no. : J. C. L. Gibson, Textbook of Syrian Semitic inscriptions III: Phoenician inscriptions (), no. .
. G. A. Cooke, Textbook of North Semitic Inscriptions (), nos. , .
. Cooke (n. ), no. ; KAI, no. ; Gibson (n. ), no. .
. P. M. Fraser, ‘‘Greek-Phoenician Bilingual Inscriptions from Rhodes,’’ ABSA 
(): .
The Phoenician Cities

zens of Phoenician cities, who were not peasants and were well enough off
to leave inscriptions, could not merely deploy both languages but could use
equivalents both for the names of deities and for their own names. They did
not have to inscribe in Phoenician as well, but chose to do so. In a small
way these inscriptions are precise examples of that Verschmelzung which some
modern scholarship has tended to deny.
A similar pattern is visible if we follow the communal life of the Phoenician cities through the period of transition. I will not try to rehearse the
complicated evidence for the dynasties of Phoenician kings whom Alexander found. For instance, the dynasty of the kings of Sidon has been reconstructed in two quite different ways down to Abdalonymus, who was in
power when Alexander arrived.29 What is clear is that some kings remained
in power after Alexander, and continued to issue Phoenician coinage in their
own names, for instance in Arados, Byblos, and Tyre (where on the coinage
of King Azemilkos continues from / to / ..).30
There was thus a certain continuity through the period of Alexander’s
conquests. But a change does seem to come in the first half of the third century. Philocles, ‘‘king of the Sidonians’’ (whose origins and career are highly
uncertain),31 is last heard of in about  ..; and the era ‘‘of the people of
Sidon’’ (above) probably starts soon after that; the era of Tyre starts from ;
and Arados appears to have lost its king by about , when a new era starts,
and the coinage makes no further reference to a king;32 at about the same
time the inventories of Delos record offerings by the ‘‘Byblioi’’ collectively,
with no king mentioned.33 So what happened? Are we to envisage a deposition of the kings and a grant or ‘‘charter’’ of the status of a Greek polis to these
cities? That some such grant could happen is suggested by the request to Antiochus IV for the Jerusalemites to be enrolled as ‘‘Antiochenes’’ in the late s
( Macc. :). A. H. M. Jones implicitly, if not quite explicitly, identified this
mid-third-century change in Phoenicia as an administrative act carried out
from above: ‘‘a thorough reorganisation of the administrative system took
. Compare E. T. Mullen, ‘‘A New Royal Sidonian Inscription,’’ BASOR  ():
, and M. Dunand, ‘‘Les rois de Sidon au temps des Pérses,’’ Mél. Univ. St. Joseph  (–
): .
. A. Lemaire, ‘‘Le monnayage de Tyr et celui dit d’Akko dans la deuxième moitié du
quatrième siècle av. J.-C.,’’ RN  (): .
. See J. Seibert, Historia  (): .
. J.-P. Rey-Coquais, Arados et sa pérée aux époques grecque, romaine et byzantine (),
–.
. E.g., Ins. Délos, no.  a. . See Ph. Bruneau, Recherches sur les cultes de Délos (),
.

The Hellenistic World and Rome
place. In the first place the Phoenician dynasties were suppressed.’’ 34 An even
more confident characterisation is given by S. K. Eddy: ‘‘The Phoenician
cities of Tyre and Sidon, a part of the Ptolemaic Empire, were reorganised as
hellenic poleis during the latter part of the third century, and they accepted
this without protest.’’ 35 But in fact it is not easy to see this process quite in
those terms, if only because it took place more or less simultaneously in areas
which were then under Ptolemaic control, such as Tyre and Sidon, and in
the Seleucid Empire, at Arados. Thus a Phoenician inscription of  ..,
from Umm El-’Amed near Tyre, dated by the th year of Ptolemy III and
the rd year of the people of Tyre.36 In the next century another Phoenician
inscription from the same site is dated to the th year of the Seleucid era
and the rd of Tyre, i.e.,  ..37 We have already seen the Phoenician
inscription from Athens dated by ‘‘the th year of the people of Sidon’’ (text
following n.  above). These eras must certainly refer to precise events; but
as to the nature of the events we have no evidence. No specific royal actions
are attested in this connection. Moreover, whatever occurred cannot have
been either the imposition or the acceptance of a fully Greek constitution.
This is clear from the simple fact that both the documents (inscriptions and
coin legends) and the actual institutions of these cities, Tyre above all, retained a mixed, Phoenician-Greek character. For instance there is a secondcentury .. inscription in Phoenician from Sidon which mentions what is
evidently a man’s official position, but it is unintelligible to us; the phrase is
‘RB ‘BR LSPṬ RB ŠNY.38 Moreover, several persons each described as ŠPṬ
(shofet: judge) are mentioned on an inscription from Tyre which is undated,
but evidently later than the period of the kings;39 we may note also a priest
(KHN) of MLK-‘ŠTRT at Umm el-’Amed near Tyre.40 As is well known,
Bickerman suggested convincingly that we should see such a judge (shofet) in
the dikastēs named Diotimus mentioned on an inscription of circa  from
Sidon (see text to n.  below). These documents, however difficult their
interpretation may be, are surely sufficient to prove that public offices identified by Phoenician names existed in some cities after the supposed grants
. A. H. M. Jones, Cities of the Eastern Roman Provinces 2 (), –.
. S. K. Eddy, The King Is Dead: Studies in the Near Eastern Resistance to Hellenism, –
 .. (), .
. KAI, no. ; cf. M. Dunand and R. Duru, Oumm el-’Amed (), , no.  (without
text); Gibson (n. ), no. .
. Cooke (n. ), no. ; KAI, no. ; cf. Dunand and Duru (n. ), , no.  (no text).
. Cooke (n. ), no. ; cf. BES , no. , for a discussion.
. Cooke (n. ), no. .
. Dunand and Duru (n. ), , no. .
The Phoenician Cities

of Greek constitutions by kings—events for which in any case there is no
direct evidence.
It was long after the commencement of the ‘‘era of the people of Tyre’’
that they did gain a privilege from a king, namely autonomia; the grant is attested by Strabo, and the coins indicate a date of / ..41 A royal grant is
also mentioned in the case of Arados, from the Seleucid king Seleucus Callinicus (– ..); but the privilege mentioned is not recognition as a
Greek city but the right of asylum (Strabo, Geog. ). However the coins
of Arados, which begin in / .., date by an era of  .. At any rate
whatever was being celebrated at Arados in the coining of tetradrachms between / and , it cannot have been Arados’ status as a purely Greek city,
since a large proportion are stamped in Phoenician letters; Greek lettering is
of course also found.42
There are as yet no inscriptions from Arados itself from the Hellenistic period. When we finally arrive at the earliest inscription from Arados
(IGLS VII, no. ), we are already in the reign of Augustus, / ..; but
the inscription is bilingual, in Phoenician and Greek. The dedicant is a gymnasiarch, but bears a Phoenician name. The deities mentioned are Hermes
—on this occasion the name is transliterated into Phoenician in the form
’RM[..]—and Herakles/Melqart. This seems also to be the latest firmly dated
inscribed text in Phoenician. There is, however, one dedicatory Phoenician
inscription from a temple in Byblos, which may be of the first century ..
(KAI, no. ). In the Roman period, so far as inscriptions reveal, the cities
of Phoenicia appear as entirely Greek. We may think for instance of the famous inscription of the Tyrian traders at Puteoli in ..  (OGIS ), with
their letter addressed to the archontes (magistrates), boulē (council), and dēmos
(people) of the city, and with the acta (proceedings or decisions) of the boulē
which followed in response. None the less, two things are apparent in the
inscriptions; the persistence of Phoenician names, and the battery of distinctive titles which the major cities claimed—‘‘hiera [holy], asylos [with the right
of asylum], autonomos, mētropolis [mother-city] of Phoenice and the cities of
Coele Syria.’’ Both features are attested for instance in the inscription put
up at Didyma (Ins. Didyma, no. ) by Tyre in honour of Julius Quadratus
of Pergamon; the ambassadors were called Marion son of Marion and Zoilos son of Bodas. The inscriptions thus both embody a claim to a glorious
past and reflect a local dispute for precedence which was to be long-lasting
even by the standards of the ancient world. For it was in .. / that
. Strabo, Geog. . For the coins BMC Phoenicia CXXV, –.
. See H. Seyrig, ‘‘Aradus et sa pérée sous les rois séleucides,’’ Syria  (): .

The Hellenistic World and Rome
Theodosius II and Valentinian III ruled that Berytus also should have the title
of metropolis. Tyre, they say, should not feel that her dignity was impaired:
‘‘Let her [Tyre] be mother of the province [mater provinciae] by the beneficium
[favour] of our ancestors, and Berytus by ours’’ (CJ , , ). ‘‘Mother of the
province’’ is a distinctive (though not unique) term, which might reflect the
Phoenician expression ’M ṢDNM—‘‘mother of the Sidonians,’’ which had
appeared on the coins of Tyre in the Seleucid period, six centuries before.
In any case, if the inscriptions of the Roman period (so far at least) hardly
reveal any Phoenician elements, the coins certainly do. It is true that it is only
in the Hellenistic period that the coins carry intelligible legends in Phoenician. But under the Roman Empire the coins of Sidon and Tyre continue
to show occasional whole words, and very frequently individual Phoenician
letters; and in the case of Tyre these persist up to the late second century ..43
That need not of course mean very much in terms of the survival of language. The letters could even have been a fossilised feature which no longer
conveyed any specific meaning. But even so, once again, they did not have to
be used. The fact that they were at least reflects a consciousness of continuity
with a city’s Phoenician past.
But was that all? The answer depends on whether we put any degree of
trust in what our literary sources tell us.
According to Josephus, Tyre possessed archives which went back to the
period of Hiram and Solomon. A number of passages in the Antiquities and
Against Apion use their evidence as support for the authenticity of the Old
Testament narrative. In most cases Josephus claims to be dependent on Hellenistic writers who had translated these documents, namely Menander of
Ephesus (Ant. ,  C. Ap. , –; Ant. , –) or Dius (Ant. , –
; C. Ap. , –). But in Ant. ,  he is more explicit about the preservation of these documents: ‘‘To this day there remain copies of these letters,
preserved not only in our books, but also by the Tyrians, so that if anyone
wished to learn the exact truth, he would, by inquiring of the public officials in charge of the Tyrian archives [ gazophylakion] find that their records
are in agreement with what we have said’’ (Loeb trans.). Of course there
are many difficulties about this. Josephus may still be repeating a Hellenistic
source; and even if a Hellenistic writer truthfully related having translated
a document preserved at Tyre, that document may itself have been a historical forgery. Moreover, the contents of the archives at Tyre are always, as
quoted, alarmingly early. No one claims to have used a continuous archive
coming down through the Persian and Hellenistic periods; the furthest we
. See BMC Phoenicia, – (Sidon); – (Tyre).
The Phoenician Cities

get is from Josephus (C. Ap. , –): a report of a Phoenician record of
the kings of Tyre down to Cyrus’ time. None the less, either what Josephus
says is just false, or it was at least believed at Tyre that they possessed records
which went back to the tenth century .. (Phoenician writing in fact went
back earlier than this). But in any case it does not matter, since what is relevant for us is the sense of continuity with a pre-Greek past. Moreover, it was
certainly believed in the Graeco-Roman world that there was an independent Phoenician historical tradition. So, in Athenaeus’ Deipnosophistae, one of
the diners, addressing Ulpian of Tyre, the main speaker, talks of ‘‘those who
write Phoenician histories, your compatriots, Sanchuniathon and Mochos’’
(A). Mochos was supposed to have been a Sidonian who had lived before the Trojan War; Posidonius alleged that he had discovered the theory
of atoms (Strabo, Geog. ); Josephus mentions him in support of biblical
reports of the longevity of the patriarchs (Ant. , ). More relevant, he was
one of three Phoenician writers whom one Laetus, an author of biographies
of philosophers, is said by Tatian (ad Gr. ) to have translated into Greek.
Sanchuniathon, from Berytus, was also said to have written in Phoenician
in the same period, and it was his work on Phoenician mythology which
Philon of Byblos claimed to have translated in the early second century. The
survival of some long extracts of this work in Eusebius’ Praeparatio Evangelica
may be due to another Phoenician, Porphyry of Tyre (whose real name, Malchus, was Semitic, ‘‘Porphyrios’’ being a pun on Malchus—Melech—king).
Nobody is going to reach agreement on the age or authenticity of the Phoenician original, if indeed there was one, as a literary work. An unprecedented
volume of modern work on Philo’s History has concurred in the view that
its entire structure is Hellenistic.44 Philo himself may have composed it de
novo; or perhaps there may have been an early Hellenistic original, a parallel to the nationalistic expositions of Berosus and Manetho. It is also worth
noting that the point that what passed as an independent Phoenician religious view-point was actually Greek had already been made by Pausanias
(, , –). At Aegium, he reports, a Sidonian argued with him, saying that
Phoenician ideas of divinity were superior to Greek ones; they said that Asclepius was air, and the son of Apollo, the Sun—Pausanias replied that this
account was no more Phoenician than Greek. But it seems beyond all question that Philo’s material does contain elements which are not derivable from
. FGrH . See for instance L. Troiani, L’opera storiografica di Filone da Byblos ();
R. A. Oden, ‘‘Philo of Byblos and Hellenistic Historiography,’’ PEQ – (): ; and
a monograph containing text, translation, and discussion, A. I. Baumgarten, The Phoenician
History of Philo of Byblos ().

The Hellenistic World and Rome
Greek traditions, and which can be brought into relation with the Phoenician divinities as they appear in the Ugaritic tablets and elsewhere. The
entire work as preserved is based on Greek conceptions; but what can categorically not be asserted is that the material which finally came through to
Eusebius, in Caesarea, was fictional, and not Phoenician at all. Among the
fragments of Philo we may note also that in talking about human sacrifice
in circumstances of extreme danger, he remarks that among the Phoenicians
an only son is still called by a term, which he transliterates as Ἰεούδ in Greek
(clearly reflecting the Semitic root for ‘‘one’’; FGrH , F b), and their
god is (still) called El. Furthermore, Eusebius, probably following Porphyry,
claims that what was reported from Sanchuniathon gained credibility ‘‘from
the names of the gods still used in the cities and villages of Phoenicia, and
the myths attached to the mysteries celebrated among them’’ (PE , , ).
It is entirely consonant with this that modern detailed studies of the Phoenician History regard it not as good evidence for an actual literary work of
the second millennium .., but as a reflection of Greek interpretations of
Phoenician religion.
It would also be valuable if we could understand how ‘‘Phoenician’’ (if at
all) Porphyry himself was; and on this I still cannot find any evidence.45 That
his name was Phoenician, and that he and his friends knew what it meant, is
hardly very significant; nor is Eunapius’ statement that he came ‘‘from Tyre,
the foremost city of the ancient Phoenicians’’ (Vit. Soph. ). But perhaps it
is, precisely in that that was how it seemed relevant to Eunapius to characterise the place: it did not apparently matter to him that since about  the
city had been a Roman colonia. We know this best of course from the other
Ulpian of Tyre, not the Greek Sophist of Athenaeus but his relative (as he
surely was), the lawyer: ut est in Syria Phoenice splendidissima Tyriorum colonia,
unde mihi origo est (as is the most magnificent colony of the Tyrians in Syria
Phoenice, from where I come, Dig. , , ). He is stretching a point, or passing over one; for when he was born, which was certainly not later than 
when the boulē (council) dismissed the complaints of the traders, and long
before /, when Diodorus son of Nithumbalus made his dedication to
Herakles/Melqart, Tyre was still a Greek city. But was it perhaps something
else as well? In Ulpian’s eyes certainly: for he goes on: ‘‘distinguished among
its neighbours, of the most remote antiquity, powerful in war, most tenacious of the treaty it struck with Rome.’’ By ‘‘of the most remote antiquity’’
he did not mean merely that it had been a Hellenistic Greek city. If he did,
. See F. Millar, ‘‘Porphyry: Ethnicity, Language and Alien Wisdom,’’ in J. Barnes and
M. T. Griffin, eds., Philosophia Togata II: Plato and Aristotle at Rome (), – ( chapter 
in the present volume).
The Phoenician Cities

there had still been in his lifetime the Phoenician letters on the city’s coins
to remind him otherwise, as well as the Phoenician names borne by his fellow citizens. Did they also still speak a language distinct both from Greek
and from Aramaic, as they certainly had when Meleager of Gadara, in about
 .., composed his well-known epigram about how to greet someone
in Aramaic, Greek, and Phoenician? 46 There is no direct evidence. But it is
Ulpian who states that an obligatio (legal obligation) could undoubtedly be
created by a question in Latin and an answer in Greek or vice versa, and who
asks whether this principle could be extended ‘‘to another language, perhaps
Punic or Assyrian’’ (ad alium [sermonem], Poenum forte vel Assyrium, Dig. , , ,
). ‘‘Assyrian’’ (Assyrius sermo) is certainly Aramaic; so perhaps Punic (Poenus
sermo) is not the Punic of Africa or Sardinia,47 but the other ancient language
of his own area, Phoenician. The coupling of the two languages then reads
quite naturally.
The Phoenician cities can therefore be taken to represent a rather special
case of Hellenisation, and it is very striking how small a part they have played
in discussions of the nature of Hellenism. They were not colonised or refounded by Alexander or his successors (the diadochi). They did not on the
whole lose their names, in the way that Josephus described as typical. There
are no clear indications that they were given Greek constitutions by the
Ptolemies or Seleucids. On the other hand they most certainly did not remain
un-Greek. They were not temple communities based on a priesthood, like
Babylon or Jerusalem. They did in some sense become Greek cities, whose
inhabitants were at home in the Hellenistic and Graeco-Roman worlds. Yet
the available evidence, scattered and inadequate though it is, surely suffices
to show that there was much continuity with their Phoenician past—in language and perhaps in institutions; certainly in their cults; probably in some
sort of literary tradition; perhaps in the preservation of archives; and certainly
in a continuous historical consciousness. This last was helped of course by
recognition from outside; we saw earlier how Carthage to the very end continued to send first-fruits to Tyre. How relevant was it that Punic continued
in use in the western Mediterranean long after the fall of Carthage and well
on into the Roman Empire? It is worth recalling also the dedication to Septimius Severus’ son Geta, put up at Lepcis Magna by the city of Tyre, which is
there called ‘‘the Septimian colony of Tyre, mother-city of Phoenicia and of
other cities’’ (Septimia Tyros Colonia, Metropolis Phoenices et aliarum civitatium)
(IRT, no. ). That must be an allusion to the Phoenician colonisation of
. Anth. Pal. , ; A. S. F. Gow and D. L. Page, The Greek Anthology: Hellenistic Epigrams
(), , no. iv.
. So F. Millar (n. ), –.

The Hellenistic World and Rome
North Africa, and is a parallel to Ulpian’s reference to the antiquity of Tyre.
A little later coins of Tyre show Dido supervising the foundation of Carthage.48 Would it be pure speculation to see in this consciousness one source
of the rash of grants of the status of colonia made to Syrian cities by Septimius
Severus from Lepcis, the husband of a lady from Emesa (which Herodian and
Heliodorus thought of as also being Phoenician)?
However, it is also relevant that the role of archaic Phoenicia in trade,
shipping, and colonisation was a well-established part of the Graeco-Roman
historical tradition.49 For instance, Pliny the Elder (NH , ) contrasts the
modest commercial activity of the present—purple fishing—with Tyre’s
former glory as a colonising city; Strabo in the Geography (–) similarly
emphasises the colonial past of Tyre, as well as the ship-building and trade
of Sidon, and the Sidonian development of astronomy and arithmetic, as a
result of the needs of navigation; while Pompeius Trogus in book  narrates
in detail the flight of Elissa (Dido) from Tyre and the foundation of Carthage.
But of course the most important contribution to Graeco-Roman culture made by Phoenicia had been the art of alphabetic writing. Everybody
(rightly) accepted that on the basis of Herodotus , , though Eupolemus, in
the second century .., asserted that writing had been invented by Moses,
who taught it to the Hebrews, after which it was transmitted by the Phoenicians to the Greeks.50 It was so much a part of the common stock of historical
knowledge that in the second century .. the Sophist Hadrianos of Tyre
could open his first performance in Athens with the famous words ‘‘Again
letters [ grammata] have come from Phoenicia’’ (Philostratus, VS , ). The
content of his proclamation was familiar, but the point could not have been
made with the same overtones by someone who was not from Phoenicia.
The transmission of the alphabet was of course only one element in the
legend of Kadmos, the son of Agenor, who came from Tyre or Sidon, and
after various adventures, including the search for his kinswoman, Europe,
founded Thebes and ruled there. The infinitely varied forms which the
legend took was studied by Ruth B. Edwards, Kadmos the Phoenician ().
It is extremely significant that this legend was subsequently integrated and
internalised in Phoenicia itself. We may not wish to use Achilles Tatius’
novel Leucippe and Cleitophon as evidence, with its description of a painting
. E.g., BMC Phoenicia, Tyre, no. ; Sylloge Nummorum Graecorum, Danish National
Museum: Phoenicia (), no. .
. For this tradition, see the exhaustive treatment by G. Bunnens, L’expansion phénicienne en Méliterranée ().
. FGrH , F. . See B. Z. Wacholder, Eupolemus ().
The Phoenician Cities

of Europe to be seen in the temple of Astarte in Sidon (, , ) or its brief
reference to the worship of Kadmos there (, , ). But the Hellenistic and
imperial coins of Tyre and Sidon which represent Kadmos and Europe are
incontrovertible evidence; on one Tyrian coin from the reign of Gallienus,
Kadmos is shown giving a papyrus roll to three Greeks.51 Equally good evidence is offered by the well-known inscription of the dikastēs or shofet ( judge;
see text to n.  above) from Sidon, Diotimus, who in the late third century .. won the chariot race at the Nemean games: ‘‘You were the first of
the citizens to bring back from Greece the glory of the chariot race to the
house of the descendants of Agenor. The sacred town of Cadmeian Thebes
also rejoices to see her mother-city made famous by her victories.’’ 52
We might well see the Hellenisation of Phoenicia, in the first instance, as
one of a number of examples from the eastern Mediterranean of a process
which began long before Alexander. In Phoenicia, as in the other cases, it
was based on long interaction with the Greek world,53 and received a further impulse from the ambitions and self-glorification of fourth-century
dynasts. Greek art had exerted a profound influence before Alexander; in
this context we may note the beautifully cut and entirely Greek reliefs of
the mid-fourth century on what is probably a throne from the sanctuary of
Eshmun/Asklepios in Sidon.54 But the most consistent evidence for Greek
influences is provided by the coinages of the Phoenician cities, beginning
in the fifth century; as Colin Kraay wrote, ‘‘The coinage of the Phoenician
cities is the most notable and elaborate example of the transplantation of this
Greek institution into an area culturally alien.’’ 55 It may seem a paradox, but
is not, that it was also these city coinages which were to continue longest
to exhibit Phoenician lettering (text to n.  above) and themes from the
Phoenician past (see above).
Fourth-century dynasts, in Phoenicia and elsewhere, certainly contributed something at least to the superficial Hellenisation of their areas. Thus,
for instance, we have Theopompus’ famous description of the luxury and
. BMC Phoenicia, no. ; Kadmos and Greeks: Tyre, no. , pl. XXXV.l.
. E. Bikerman, ‘‘Sur une inscription grecque de Sidon,’’ Mélanges Dussaud I (), .
Translated by M. M. Austin, in The Hellenistic World (), no. .
. See now P. J. Riis, ‘‘Griechen in Phönizien,’’ in Niemeyer (n. ),  (for the archaic
period); J. N. Coldstream, ‘‘Greeks and Phoenicians in the Aegean,’’ in Niemeyer (n. ), .
. E. Will, ‘‘Un nouveau monument de l’art grec en Phénicie: la ‘tribune’ du sanctuaire
d’Echmoun à Sidon,’’ BCH  (): .
. C. M. Kraay, Archaic and Classical Greek Coins (), –, on p. . Cf. M. H.
Crawford, La moneta in Grecia e a Roma (), –, suggesting Persian influence on the
pattern of minting in Tyre and Sidon.

The Hellenistic World and Rome
display of Straton of Sidon, his hiring of entertainers from Greece, and his
rivalry with Nicocles of Salamis (Athenaeus, Deipn. ). It was either this
Straton or the one whom Alexander found in power under whom merchants
from Tyre dedicated statues representing Tyre and Sidon to Delian Apollo,
commemorating this with a bilingual inscription.56 The impulse given to
Hellenisation in Caria by the dynast is equally the main theme of Simon
Hornblower’s Mausolus (); and the dynastic monuments, the bi- and trilingual inscriptions, and the coins of fifth- and fourth-century Lycia provide
ample evidence of comparable influences there.57 We are thus presented with
the agreeable paradox that the most clear and convincing cases of Verschmelzung in Droysen’s sense from the Greek East with which our evidence now
presents us were ones which began before Alexander, and where it remains
quite unclear how much their cultural evolution owed to the Macedonian
conquest.
In the case of Phoenicia, however, there was perhaps more to it than
that. Firstly, the Phoenician cities already bore at least some resemblance to
Greek city-states; it is not easy to say what if any significant social changes
their (partial) evolution into Greek poleis will have necessitated. Secondly,
and more important, when the Phoenicians began to explore the storehouse
of Greek culture, they could find, among other things, themselves, already
credited with creative roles—not all of which, as it happens, were purely
legendary. If some aspects were just legend, like the story of Kadmos, what
is clear is that the Phoenicians adopted it (perhaps, like the legend of Aeneas
in Italy, very early) and made it their own. In doing so they acquired both
an extra past and a reinforcement of their historical identity; and they also
simultaneously gained acceptance as being in some sense Greeks. It is relevant to observe that the Romans, admitted to the Isthmian games in  ..,
probably a little before Diotimus of Sidon won the chariot race at Nemea,
earned a similar acceptance on similar grounds. The famous inscription of
the s from Lampsacus shows that the city appealed to the Romans as ‘‘relatives,’’ on the basis of their descent from Troy.58
That was also, it need hardly be said, the period of the birth of Latin literature. We might well wonder why it was in Rome and not in Phoenicia that
there evolved, entirely without the aid of a conquering Macedonian state,
the only literary culture which really was a ‘‘fusion’’ in Droysen’s sense.
. CIS ., no. : Ins. Délos, no. .
. See W. A. P. Childs, ‘‘Lycian Relations with Persians and Greeks in the Fifth and
Fourth Centuries Reexamined,’’ Anat. Stud.  (): .
. Syll.3 ; I. K. Lampsakos (), no. . Translated by Austin (n. ), no. .
 
Hellenistic History in a
Near Eastern Perspective:
The Book of Daniel *
The Background
As with so many aspects of Hellenistic history, it is best to begin with the
words of Polybius: ‘‘I shall bring the whole narrative of events to a conclusion, narrating finally the expedition of Antiochus Epiphanes against Egypt,
the war with Perseus and the abolition of the Macedonian monarchy.’’ Polybius is here developing his ‘‘second introduction,’’ in the first few chapters of
book , in which he goes on to explain that, after all, he has changed his mind
about the terminal date. It was not now to be, in modern terms,  .., but
. For his secondary purpose would be to examine how the Romans had
used the power that they had gained, and the new terminal point would be
the destruction of Carthage and the disaster which happened in Greece in
the same year: ‘‘the withdrawal of the Macedonians from their alliance with
Rome and that of the Lacedaemonians from the Achaean League, and hereupon the beginning and end of the general calamity that overtook Greece.’’ 1
Artistically, it might be argued that it would have been better to keep to
the original plan, to cover the period (/– ..) in which ‘‘in not quite
fifty-three years all parts of the inhabited world [oikoumenē] fell under the
single rule [archē] of the Romans.’’ 2 The period had been one of rapid evo* First published in P. Cartledge, P. Garnsey, and E. Gruen, eds., Hellenistic Constructs: Essays
in Culture, History, and Historiography (Berkeley, ), –. I was very grateful to Martin
Goodman,Tessa Rajak, and Geza Vermes for various forms of help, guidance, and correction
in the preparation of this paper.
. The ‘‘second introduction’’ occupies bk. , –. The passages quoted, from the Loeb
translation, are , , –, and , , .
. Polybius , .


The Hellenistic World and Rome
lution, and the theme had coherence and dramatic force. As Peter Derow
showed some years ago, what Polybius meant by archē was not the formation
of Roman provinces, but the developing capacity to defeat some peoples or
rulers, and to give instructions to others.3 Both aspects of Roman domination
are perfectly exemplified in book , which would have formed the original
conclusion: the defeat of Perseus at the battle of Pydna, and the famous scene
outside Alexandria when the Roman commander, Popilius Laenas, ordered
the Seleucid king Antiochus IV Epiphanes to abandon the siege of the city
and go home.4 Polybius could not of course know that, in historiographical terms, this moment had already acquired a significance even greater than
it manifestly had as a historical event. For the author of Daniel had already
made it the latest event to be recorded in the Hebrew Bible (:): ‘‘For the
ships of Chittim will come against him; therefore he will be disheartened
and return.’’
The connections and contrasts between Polybius and Daniel are far more
profound, however, than this accidental conjunction. In Polybius’ conception of history, the Romans were seen as taking their place in a succession
of world rulers: first the Persians, then the Spartans, then the Macedonians,
and now the Romans (, ). Moreover, it was in his originally planned final
book, , that Polybius included an extended quotation of the reflections of
Demetrius of Phalerum, meditating on the destruction of the Persian Empire
by Alexander:
For if you consider not countless years or many generations, but merely
the last fifty years, you will read in them the cruelty of Fortune. I ask
you, do you think that fifty years ago either the Persians and the Persian king or the Macedonians and the king of Macedon, if some god
had foretold the future to them, would ever have believed that at the
time when we live, the very name of the Persians would have perished
utterly—the Persians who were masters of almost the whole world—
and that the Macedonians, whose name was formerly almost unknown,
would now be the lords of it all? 5
The idea that the history of mere ‘‘events’’ was something superficial or insignificant would have amazed Demetrius and his contemporaries in the Greek
world, just as it would have surprised later observers in the Near East who
reflected on the devastating impact of Alexander’s conquests. One such was
. P. Derow, ‘‘Polybius, Rome, and the East,’’ JRS  (): –.
. Polybius , .
. Polybius , , –, Loeb trans.
The Book of Daniel

the author of  Maccabees, writing (almost certainly) some time towards the
end of the second century .. His work opens as follows:
And it came to pass after the victory of Alexander the son of Philip
the Macedonian, who came from the land of Chittim, that he smote
Darius, king of the Persians and Medes, and ruled in his stead, beginning in Greece. He waged many wars and captured strongholds and
slew (the) kings of the land. He pressed forward to the ends of the earth
and took spoils from a multitude of peoples, and the earth was silent
before him, and he was exalted and his heart filled with pride.6
We can assume that in Judaea itself Alexander’s passage, from Tyre down the
coast to Egypt, will have made a corresponding impact. The only detailed
narrative reflection of it from a Jewish perspective which we have, however,
is a legend, or perhaps better a religious historical novella, which Josephus
incorporated in book  of his Antiquities.7
We shall return briefly later to the significance of Josephus’ use here of
Daniel, which in fact had not yet been written at the time described in Ant.
. It is more important to stress at this point that the still developing corpus
of biblical works did incorporate a reasonably complete historical narrative,
or series of narratives, relating the impact on the Jewish community of a succession of Near Eastern empires, going back through the Persian Empire of
the Achaemenids to the Babylonian and Assyrian empires. To go no further
back, the book of Kings contained a historical narrative from the end of King
David’s reign to the Babylonian Captivity. The two books of Chronicles retold the story from the time of Saul onwards, continuing to a rather brief
account of the Captivity, and ending triumphantly with the victory of Cyrus
over the Babylonian dynasty ( ..) and the return from the Captivity:
In the first year of Cyrus king of Persia (KWRŠ MLK PRS)—to fulfil the word of Yahweh through Jeremiah—Yahweh roused the spirit
of Cyrus king of Persia to issue a proclamation and to have it pub.  Macc. :. The sequence of tenses in this section is not clear, and the translation
is hypothetical. For a basic discussion of  Macc., generally thought to be a Greek version
of a Hebrew or Aramaic original, see Schürer, Vermes, and Millar, History III., –. A
relatively early date of writing, of around  .., is argued by S. Schwartz, ‘‘Israel and
the Nations Roundabout: I Maccabees and the Hasmonean Expansion,’’ Journ. Jew. Stud. 
(): –.
. Josephus, Ant. , –; Loeb trans. of , . See, e.g., A. D. Momigliano, ‘‘Flavius
Josephus and Alexander’s Visit to Jerusalem,’’ Athenaeum  (): – Settimo contributo alla storia degli studi classici e del mondo antico (), –.

The Hellenistic World and Rome
licly displayed throughout his kingdom: ‘‘Cyrus king of Persia says this
‘Yahweh, the God of Heaven, has given me all the kingdoms of the
earth and has appointed me to build him a Temple in Jerusalem, which
is in Judah. Whoever there is among you of all his people, may his God
be with him! Let him go up.’ ’’ 8
As is evident, Chronicles could not have been written as it stands before the
Achaemenid period, and it is sometimes argued that it was written, or edited,
as late as the early Hellenistic period. The same must apply to the books
of Ezra (which begins with a slightly extended version of the same proclamation by Cyrus) and of Nehemiah, which between them relate, within a
structure which is chronologically confused and not mutually related as between the two books, the restoration of the Temple and the re-establishment
of Jewish worship and observances under the Achaemenids, thus taking the
story to some (now indeterminable) point in the fifth century.9
An alternative version of the same story (but equally confused in its conception of the sequence of Persian kings) was constructed at some time in
the Hellenistic period, using material found also in  Chronicles, Ezra, and
Nehemiah, and provides a continuous narrative going from Josiah to Ezra.
This is the important and highly puzzling text called  Esdras, which may
have been composed originally in Hebrew or Aramaic, but certainly existed
in Greek by the time of Josephus, who used it for his Antiquities.10
Its significance in this context is simply as another sign of awareness within
Jewish culture of the national tradition as a historical narrative, and as one
whose course had been dependent on the attitudes and actions of a succession of external rulers. Such an awareness was quite compatible with a lack of
complete clarity as to which ruler had succeeded which, or even as to which
empire had been which. It was this sometimes confused impression of a sequence of empires and rulers which provided the framework for a group of
three religiously inspired historical novels, one of which, Esther, was to be
incorporated in the Hebrew Bible, while Tobit and Judith appeared in the
Christian Bible in Greek, which we refer to for convenience as the Septuagint.11 (In consequence, all three appear in Roman Catholic Bibles, but only
.  Chron. :–, trans. New Jerusalem Bible.
. See O. Eissfeldt, The Old Testament (), – (Chronicles); – (Ezra and
Nehemiah); H. G. M. Williamson, The New Century Bible Commentary:  and  Chronicles
(); and by the same author, World Biblical Commentary XVI: Ezra, Nehemiah ().
. For a detailed survey of the problems, see Schürer, Vermes, and Millar, History III.,
–.
. See the fundamental article by M. Hengel, to which I am extremely indebted, ‘‘Die
The Book of Daniel

Esther in Protestant ones.) Tobit is set in the Assyrian Empire of the eighth
to seventh century. The hero is introduced as one who ‘‘in the days of Shalmaneser king of the Assyrian empire was carried away captive out of Thisbe
which is on the right hand of Kedesh Naphtali in Upper Galilee.’’ Thenceforward he lived in Nineveh, in the reigns of Shalmaneser (V), Sennacherib,
and Esarhaddon (the sequence is historical, if incomplete, the respective dates
being –, –, and – ..). The full text of Tobit survives
only in Greek, but it is of crucial importance that fragments in Aramaic,
with one in Hebrew, have been found in Qumran.12 The work is thus almost
certainly a Semitic-language composition of the early Hellenistic period.
Judith, by contrast, confuses the Assyrian and Babylonian empires: ‘‘It was
in the twelfth year of Nebuchadnezzar who reigned over the Assyrians in the
great city of Nineveh.’’ The historical Nebuchadnezzar was of course a Babylonian king (see below). There is no trace of the (probable) Semitic original
of the Greek text of Judith; and only general arguments from the religious
and nationalistic tone of the story of Judith and Holofernes might perhaps
suggest that it could belong in the Maccabean period.13
As for Esther, if we ignore an inept introductory paragraph added to the
Greek version, which confuses the Persian and Babylonian empires (as does
indeed a reference to Esther’s father in :–), it is firmly set in the Persian Empire, in the reign of a king ’ḤŠWRWŠ (Ahasuerus-Artaxerxes), who
ruled from Susa over an empire which stretched from India to Kush, and
who had an army of Persians and Medes (ḤYL PRS WMDY). The whole
story of Esther is retold in detail by Josephus in the Antiquities, where it is
set in the fifth century.14 As a literary work, it was certainly in existence by
the second century .. at the latest. A note at the end of the Greek version
claims that one Dositheos had ‘‘brought’’ the letter concerning Purim (described in :–) in the fourth year of Ptolemy and Cleopatra, after it had
been translated by Lysimachus, a Jerusalemite. If this reference is authentic,
and not merely a piece of pseudo-historical colouring, the date referred to
Septuaginta als ‘christliche Schriftensammlung’: Ihre Vorgeschichte und das Problem ihres
Kanons,’’ in M. Hengel and A. M. Schwemer, eds., Die Septuaginta zwischen Judentum und
Christentum (), –.
. For Tobit, see R. H. Charles, Apocrypha and Pseudepigrapha of the Old Testament I (),
–; P. Deselaers, Das Buch Tobit (); Schürer, Vermes, and Millar, History III., –.
For Qumran fragments Q–, see now J. A. Fitzmyer, Qumran Cave , XIV: Parabiblical
Texts, part , DJD XIX (), –.
. See Charles (n. ), –; Eissfeldt, The Old Testament, –; Schürer, Vermes,
and Millar, History III., –.
. Josephus, Ant. , –.

The Hellenistic World and Rome
is / .. There are no known fragments from Qumran, making this the
only book of the Hebrew Bible not represented there.15
The establishment of the feast of Purim thus was, or was made into, the
central point of this vivid court narrative, and this feature is reflected clearly
in  Maccabees, surely written in the second half of the second century ..
This book ends with the defeat of the Seleucid general Nicanor in  ..,
and the decision to celebrate the ‘‘day of Nicanor’’ on the thirteenth of Adar,
on the day before ‘‘the day of Mordechai’’: ‘‘the thirteenth day of the twelfth
month—it is called Adar in the Syrian language—one day before the day
of Mordechai.’’ 16 Purim was thus already established as a festival and was already associated with the story of Esther, Haman, and Mordechai. Esther is
thus (perhaps) roughly contemporary with Daniel, or might be as much as
a couple of centuries earlier. It shares with Daniel the use of history, and
the deployment of an alien court as a stage for the demonstration of Jewish
piety.
The Book of Daniel
These varied works, for all the problems of date, authorship, intention, and
original language which they present, are enough to set the framework for
Daniel, and to suggest the way in which the successive Near Eastern empires, and the relationship of the Jewish community to them, could be used
in Jewish historical and semi-historical works of the late Achaemenid and
earlier Hellenistic period, in ways which ranged from sequential narratives
of events to colourful and improving historical novels.
As will be obvious, this paper will make no attempt even to indicate all
the problems which surround Daniel: the question, for instance, of whether
there had been earlier versions before the (indisputable) redaction of the canonical Hebrew and Aramaic text in the s ..; the meanings intended
by the anonymous author to be attached to the symbols incorporated in the
dreams and visions which it recounts; whether even after the s variant
‘‘Daniel’’ texts were in circulation; the dates and origins of the two established Greek versions; and the interpretation of the text in the New Testament, in Josephus, by Christian commentators, such as Hippolytus and
Jerome, or by Daniel’s brilliant pagan commentator, Porphyry. All that is intended here is to set out the most elementary facts, and then to present an
. For Esther, see Eissfeldt, The Old Testament, –, and – (the additions). For
the additions, see also Schürer, Vermes, and Millar, History III., –.
.  Macc. :.
The Book of Daniel

analysis of the structure of the text that will show just how detailed a use it
makes of Near Eastern history from the sixth century up to the s ..
As for the date and context of the canonical Daniel as we have it, no serious commentator would now question Porphyry’s demonstration that the
work belongs in the s under Antiochus IV Epiphanes, and that up to and
including that point the prophecies in it are pseudo-prophecies, relating and
giving meaning to events which had already occurred. At the time when
the author was completing the work, the imposition of the ‘‘abomination of
desolation’’ in  had occurred, and the restoration of the Temple cult in
 had not.17 The ‘‘reception’’ of Daniel is a topic which could take many
volumes and is integral to early Christianity, in particular to the notion of
Jesus as a ‘‘Son of Man’’ (see briefly text to n.  below). But it may be noted
that stories which form part of ‘‘our’’ Daniel are alluded to in  Maccabees,
which (as above) dates to the later second century .., or (less probably)
to the early first. In the ‘‘testament of Mattathias’’ in  Macc. , recounting
historical examples of pious Jews, there appear (:) Ananias, Azarias, and
Misael saved from the furnace (cf. Daniel ); and (:) Daniel saved from
the lion (cf. Daniel ). Similar allusions appear in  Maccabees (perhaps of
the first century ..) and  Maccabees (probably of the first century ..).18
Much more important still is the fact that actual fragments of the canonical Daniel (as well as other Daniel-like prophetic fragments) are known
from Qumran and would thus count among the textual attestations from
the ancient world which are closest in time to the original composition.19
. For the essential treatments and bibliography, see R. H. Charles, A Critical and Exegetical Commentary on the Book of Daniel (); L. F. Hartman and A. A. De Lella, The Book
of Daniel, Anchor Bible  (); Schürer, Vermes, and Millar, History III., –; and now
the massive commentary by J. J. Collins, Daniel: A Commentary on the Book of Daniel ().
Note also E. Bickerman, Four Strange Books of the Bible: Jonah, Daniel, Koheleth, Esther (),
and A. S. van der Woude, ed., The Book of Daniel in the Light of New Findings (). In this
latter collection, note esp. M. Delcor, ‘‘L’histoire selon le livre de Daniel, notamment au
chapitre ,’’ –, and C. C. Caragounis, ‘‘History and Supra-history: Daniel and the Four
Empires,’’ –.
.  Macc.  (the prayer of Eleazar): section  (three companions in Babylonia saved
from furnace); and  (Daniel saved from lion). See J. H. Charlesworth, The Old Testament
Pseudepigrapha II (), . Cf.  Macc. :; :; :; :. See E. Bickerman, ‘‘The
Date of Fourth Maccabees,’’ in Studies in Jewish and Christian History I (), ff., and
Charlesworth, The Old Testament Pseudepigrapha II, –.
. See E. Ulrich, ‘‘Daniel Manuscripts from Qumran I,’’ BASOR  (): –;
‘‘II,’’ BASOR  (): –. Final publication by the same author in Qumran Cave  XI,
DJD XVI (), Q–Q ( QDan a–e ); see also Qumran Cave , DJD I (), appen-

The Hellenistic World and Rome
The earliest of the Daniel fragments from Qumran Cave IV covers sections
of chapters –, and is thought to date from the late second century ..,
thus only some half a century after the composition of these chapters. The
fragments also confirm that the shift from Hebrew to Aramaic and back
characteristic of the canonical Daniel (:b–: is in Aramaic, the rest in
Hebrew) was already present in the texts circulating in the late Hellenistic
period. At some point which remains open to debate two different Greek
translations were made, one of which came to be that which formed part of
Christian codices of the Bible and is therefore labelled as the Septuagint text—
quite unhistorically, since the legendary account of the translation carried
out by seventy-two elders under Ptolemy Philadelphus related only to the
five books of Moses, and at that time (the third century ..) the canonical
Daniel had not been written. The other is that ascribed to Theodotion, alleged to have worked in the second century .. But a text which was at any
rate close to his is quoted by writers of the first century .. In particular,
a ‘‘pre-Theodotion’’ version seems to lie behind some passages in the New
Testament.20 But it is futile to try to see the earliest history of the translation of Daniel into Greek in terms of specific and distinct ‘‘versions’’ represented in later manuscripts. For by far the most important known user of
the Greek Daniel is Josephus, in book  of the Antiquities, written in Rome
in the s ..21 His paraphrase of Daniel incorporates readings which are
shared with either the ‘‘Septuagint’’ text or the ‘‘Theodotion’’ one, or with
neither.22
It is in fact not unlikely that Daniel had been translated into Greek, perdix, nos. – ( QDan a–b ), and Les petites grottes de Qumran, DJD III (), – (Q
PapDan). Note in particular the Aramaic ‘‘son of God’’ fragment, discussed by G. Vermes
in ‘‘Qumran Forum Miscellanea I,’’ Journ. Jew. Stud.  (): –, on –; translated
in G. Vermes, The Dead Sea Scrolls in English 4 (), –.
. For these questions, which I do not attempt to go into further, see M. Harl, G. Dorival, and O. Munnich, La Bible grecque des Septante (), and the important article by
M. Hengel, ‘‘Die Septuaginta’’ (n. ). I was very grateful to Tessa Rajak for allowing me
to see the text of her unpublished Grinfield Lecture on the Septuagint, Oxford University, : ‘‘The Jewish Reception of the LXX in the First and Second Centuries ..: Some
Thoughts on Josephus and LXX Daniel.’’
. Ant. , –; , –. See G. Vermes, ‘‘Josephus’ Treatment of the Book of
Daniel,’’ Journ. Jew. Stud.  (): –, and S. Mason, ‘‘Josephus, Daniel and the Flavian
House,’’ in F. Parente and J. Sievers, eds., Josephus and the History of the Greco-Roman Period
(), –.
. See J. Ziegler, ed., Susanna, Daniel: Bel et Draco (Göttingen Septuagint XVI., ),
. As does Rahlfs in his standard edition, Ziegler prints both texts.
The Book of Daniel

haps more than once, as early as the later Hellenistic period, as we know was
the case with Esther (text to nn. – above) and with the Wisdom of Ben
Sira (Ecclesiasticus).23 But all we can be certain of is that it was in circulation
in Hebrew and Aramaic by the end of the second century .. and in Greek
also by the first century ..
Josephus, as we saw earlier in relation to Alexander’s legendary visit to
Jerusalem (text to n.  above), accepted ‘‘Daniel’’ as a historical personage of the sixth century .. living under the Babylonian Captivity, and
hence treated his dreams and visions as genuinely prophetic. He thus equally
took the references in Daniel to Antiochus Epiphanes as prophetic, and also
understood the prophetic element as including the Roman Empire. Whether
Josephus interpreted Daniel as specifically foretelling the destruction of the
Temple in ..  is, unfortunately, unclear, owing to a problem in the text.
According to a possible reconstruction of his text, Josephus wrote as follows:
And there would arise from their number a certain king who would
make war on the Jewish nation and their laws, deprive them of the form
of government based on these laws, spoil the temple and prevent the
sacrifices from being offered for three years. And these misfortunes our
nation did in fact come to experience under Antiochus Epiphanes, just
as Daniel many years before saw and wrote that they would happen. In
the same manner Daniel also wrote about the empire of the Romans
and that [ Jerusalem would be taken by them and] the temple would be
laid waste.24
Josephus’ recognition that Daniel did indeed speak about Antiochus Epiphanes is however quite unambiguous, and he reverts to this point (which is not
in his main source,  Maccabees) when he comes to relate Antiochus’ desecration of the Temple in  ..: ‘‘Now the desecration of the temple came
about in accordance with the prophecy of Daniel, which had been made four
hundred and eight years before; for he had revealed that the Macedonians
would destroy it.’’ 25 He thus supplied nearly all the evidence for Porphyry’s
correct deduction that this prophecy was a pseudo-prophecy, but of course
without drawing this conclusion himself.26
. See the Wisdom of Ben Sira, prol. –. The year is  ..
. Josephus, Ant. , –, Loeb trans. The words bracketed are supplied only from
John Chrysostom, Adv. Iudaeos ,  (Migne, PG XLVIII, col. ). See R. Wilken, John Chrysostom and the Jews (), esp. –.
. Ant. , , Loeb trans.
. Porphyry’s argument is preserved in Jerome’s Commentary on Daniel, Corpus Christ.

The Hellenistic World and Rome
Daniel: Narrative Structure
It was a pseudo-prophecy, however, and Daniel does preserve for us a unique
representation, written in the mid-Hellenistic period, of how Jewish piety
and religious knowledge had been deployed under a series of foreign empires, Babylonian, Persian, and Macedonian, over a period of four hundred
years. The following tabulation makes no attempt to go into interpretative
details, but is concerned to do two things: to make clear the literary structure of the work, which means primarily the relationship between ‘‘authorial
voice’’ (first person and third person), narrative, and interpretation; and to
set out summarily the real identities and dates of the rulers to whom the
work refers.27
It will be seen that in form and character the work divides into four distinct sections. In section (), covering chapters –, a series of narratives, in
the third person, with no identified narrator, and designed to demonstrate
the operation of Jewish piety, are interspersed with dreams and portents external to Daniel, which are then provided with interpretations by him. In
section (), covering chapters –—or perhaps better chapters –, with
chapters – forming a sort of transition to ()—Daniel himself has very
detailed prophetic dreams, occasionally described in the third person (:;
:), but related by him in the first person, whose interpretation, in even
greater detail, is supplied first by an anonymous figure to whom he turns
(:), and then by Gabriel (:). Chapters – move on to prayers by
Daniel and to exchanges between Gabriel and Daniel, both related by Daniel
in the first person. Section () covers :–, and contains a continuous series
of historical prophecies addressed to Daniel by an angelic being who is not
specifically named (:: W‘TH ’MT ’GYD LK / καὶ νῦν ἀλήθειαν ἀναγγελῶ
σοι [Theodotion]—‘‘and now I will declare to you the truth’’). The literary
structure of this section is quite different, in that there is no division between
the material and its interpretation. The narrative, interspersed with interpretation and comment by the same speaker, covers nearly four centuries,
from the beginning of the Achaemenid period to the middle of the reign of
Antiochus Epiphanes.
Lat. LXXVA, ; the relevant passage is reproduced and translated in Stern, Greek and Latin
Authors II, no. A.
. For these identifications, and the historical framework in general, see Amélie Kuhrt,
The Ancient Near East, c. – .. I–II (). Volume II covers, among other things,
the Neo-Assyrian Empire (chap. ), Babylonia, ca. – (chap. ), and the Achaemenid
Empire (chap. ).

The Book of Daniel
Finally, there is section (), covering : to the end of chapter , the
only truly prophetic part of the whole work, which looks forward, in very
significant style, to events which really had not yet occurred (and never did).
In literary structure, however, this section is continuous with () and is also
placed in the mouth of the same angelic being.
Section ()
Chapter Narrative; dreams and visions

Capture of Jerusalem by Nebuchadnezzar (II) of Babylon (– ..).
Daniel, Hananiah, Mishael, and
Azariah selected for court service.
They refuse royal rations. This
continues to ‘‘first year of Cyrus’’
( ..).

Year ‘‘’’ of Nebuchadnezzar.
His dream.

Nebuchadnezzar sets up golden
image. Hananiah (Shedrach), Mishael
(Meshach), and Azariah (Abednego)
refuse to worship. Placed in furnace,
saved. Nebuchadnezzar orders all to
respect their God.

Nebuchadnezzar’s dream of tree
(related in first person).
Interpretations
Daniel interprets dream:
image of gold, silver, brass,
iron; indicates four successive
kingdoms, to be followed by
eternal kingdom.
Daniel interprets: dream
portends Nebuchadnezzar’s
expulsion from kingdom, to
live in fields.
Nebuchadnezzar is expelled and then
restored, and worships God (related
in third, then first, person).

Feast of Belshazzar, ‘‘son’’ of
Nebuchadnezzar ( Bel-shar-uṣur,
ruling in Babylon as ‘‘royal prince’’ in
the absence of his father, Nabonidus
[– ..]). Vessels from Temple
used. Writing on the wall.

The Hellenistic World and Rome
Section () (continued )
Chapter Narrative; dreams and visions
Interpretations
Daniel interprets writing:
kingdom to be given to
Medes and Persians
(MDY WPRS).
Belshazzar killed. Kingdom taken by
‘‘Darius the Mede’’ (in fact by Cyrus,
 ..).

Darius forbids prayer to man or god,
except himself. Daniel prays all the
same. Thrown to lions, saved. Darius
orders observance of Daniel’s God.
Daniel prospers in reign of Darius
(I?) (– ..) and Cyrus
(/ ..).
Section ()

Daniel’s dream of four beasts; th
beast with  horns, and th which
destroys  of others. Day of judgement. Beast (th?) slain. Reign of
one ‘‘like a son of man’’ (KBR ’NŠ,
:).
Interpretation given to Daniel
by unnamed person (:).
Four kings/kingdoms. 
kings of th kingdom. Persecution by th. Dominion to
be given to ‘‘the people of the
saints of the most high’’
(‘MQDYŠY ‘LYWNYN,
:).

Daniel’s dream (in ‘‘year  of
Belshazzar’’). Ram, defeated by goat
from west. His horn broken, and
replaced by four horns. From one
of them comes little horn, who
magnifies himself, and destroys daily
sacrifice (TMYD, :) and sanctuary.
Interpretation given to
Daniel, by Gabriel (named in
:): ram is king of Media
and Persia (MDY WPRS);
goat is king of Greece (MLK

The Book of Daniel
Section () (continued )
Chapter Narrative; dreams and visions
Interpretations
YWN) (Alexander, –
 ..). Four kingdoms.
Then persecuting king
(Antiochus Epiphanes,
– ..).

Daniel’s vision (in year  of Darius
‘‘son of Ahasuerus’’ (DRYWŠ BN
’ḤŠWRWŠ) ( ..? Or Darius II,
– ..?); he understands
Jeremiah’s prediction of -year
desolation of Jerusalem. Daniel’s
prayer. Gabriel appears to Daniel,
predicts cessation of sacrifice and
offering, and imposition of
‘‘abomination of desolation’’
(ŠQWṢYM MŠMM / βδέλυγμα
τῶν ἐρημώσεων, :).

Year  of Cyrus king of Persia
(KWRŠ MLK PRS) ( ..?).
Daniel’s vision of angel (not named),
and then of (different?) person ‘‘like
a son of man’’ (KDMWT BNY ’DM,
:; cf. :). Latter announces
prophecy, ‘‘what is written in the
book of truth’’ (BKTB ’MT, :).
Section ()
Chapter
Verse

–




Narrative
The same person (angel?) prophesies course of events from
Darius ‘‘the Mede’’ (or Cyrus, as in LXX and ‘‘Theodotion’’). Three more kings in Persia (LPRS). Fourth
(Xerxes, – ..?) will attack ‘‘all the kingdom of
Greece’’ (HKL ’T MLKWT YWN).
Mighty king will arise (Alexander).
Kingdom divided into four (early Hellenistic monarchies).
Strength of ‘‘king of south’’ (MLK HNGB) (Ptolemy I,
– ..).
Marriage of daughter of ‘‘king of south’’ to ‘‘king of north’’
(MLK HṢPWN). Marriage breaks down. (Berenice,
daughter of Ptolemy II Philadelphus, marries Antiochus II
Theos, c.  ..)

The Hellenistic World and Rome
Section () (continued )
Chapter
Verse
–
–
–




–
Narrative
Invasion by ‘‘king of south.’’ Spoils carried to Egypt
(MṢRYM). (Ptolemy III Euergetes, – .., and Third
Syrian War, – ..; for spoils, cf. Adoulis inscription,
 .., OGIS ; Canopus decree, OGIS .)
Victory of ‘‘king of south’’ over ‘‘king of north.’’
(Ptolemy IV defeats Antiochus III at Raphia,  ..)
Conquests by ‘‘king of north.’’ (Conquest of Coele Syria and
Palestine by Antiochus III, /– ..)
Marriage alliance between kings. (Cleopatra, daughter of
Antiochus III, marries Ptolemy V Epiphanes, / ..)
(King of north) turns to isles. Frustrated. ([?] Antiochus III,
in Asia Minor, Greece, conflict with Rome, defeat at
Magnesia, treaty of Apamea, – ..)
Return; death. (Death of Antiochus III in Elymais,  ..)
His successor, death of. (Seleucus IV Philopator,
– ..)
‘‘King of north’’ invades kingdom of south, returns, action
against holy covenant. Second invasion, stopped by ‘‘ships of
Kittim’’ (ṢYYM KTYM, , ). Action against sanctuary,
suppression of daily sacrifice (TMYD), imposition of the
‘‘abomination of desolation’’ (ŠQWṢ MŠMM / βδέλυγμα
ἐρημώσεων). Resistance begins. [Antiochus IV Epiphanes,
– .. Invasion of Egypt, / .. Second
invasion,  .., stopped by Roman emissary, Popilius
Laenas. Suppression of Temple cult,  .. Maccabean
revolt begins.]
Section ()

–
–

–

Prophecies of prospective last part of Antiochus’ reign.
Prophecy of appearance of angel Michael, delivery of Israel.
Resurrection and Day of Judgement.
End of vision.
Daniel reports final vision and message.
From cessation of daily sacrifice (TMYD) and imposition of
‘‘abomination of desolation’’ (ŠQWṢ ŠMM / τὸ βδέλυγμα
[τῆς] ἐρημώσεως) for , days.
The Book of Daniel

Conclusion: Daniel and History
Many aspects of the text of Daniel, heavily loaded with often enigmatic symbols and with religious meaning, have an important place in the history of
Judaism and Christianity. It is only necessary to note, firstly, the earliest unambiguous appearance in Jewish literature of the notions of a day of judgement and of the resurrection of the virtuous (:: ‘‘Of those who lie sleeping
in the dust of the earth many will awake, some to everlasting life, some to
shame and everlasting disgrace’’).28 Secondly, there is the (highly enigmatic)
symbolism of the one ‘‘like a son of man’’ (KBR ’NŠ) seen ‘‘coming on the
clouds of heaven’’ (:–), whose relevance to New Testament imagery is
obvious.29
My concern here is however only to stress, firstly, how closely integrated
the various sections of Daniel are one with another, and to suggest how profoundly they are related to the Maccabean crisis. Cyrus is mentioned already
in :, and the Medes and Persians appear first in :, while the Hellenistic
monarchies and Antiochus Epiphanes are mentioned already in chapters –
. In chapter  we have the first of three appearances (:; :; :) of
the suppression of the daily sacrifice in the Temple (TMYD), and the imposition of the ‘‘abomination of desolation.’’ Equally important, even the earlier
part of the work is dominated by concerns over the personal observation of
Judaism, concerns which it is tempting to interpret as reflecting the strains
imposed by the persecution under Antiochus from  .. onwards, which
was noteworthy for being directed not only at the Temple cult, but at the
private observance of Judaism.30 Hence there are references to the avoidance of unclean food (chapter ); the worship of pagan gods (chapter ); and
the offering of prayers to deities, or supposed deities (chapter ). In spite of
the drastic successive shifts in ‘‘authorial voice,’’ literary form, and narrative
structure which characterise this quite brief work, it may be suggested that in
its conceptions and concerns it can be seen as a unity, and one which reflects
the great crisis of the s.
. See G. W. E. Nickelsburg, Resurrection, Immortality and Eternal Life in Intertestamental
Judaism (); E. P. Sanders, Judaism: Practice and Belief  ...– .. (), –.
. See, e.g., M. Black, An Aramaic Approach to Gospels and Acts 3 (), app. E (G. Vermes); G. Vermes, Jesus the Jew (), –; Schürer, Vermes, and Millar, History II, –;
C. Rowland, The Open Heaven: A Study of Apocalyptic in Judaism and Early Christianity ();
J. Ashton, Understanding the Fourth Gospel (), –.
. For my view of this, see ‘‘The Background to the Maccabean Revolution: Reflections
on Martin Hengel’s ‘Judaism and Hellenism,’ ’’ Journ. Jew. Stud.  (): – ( chapter 
of the present volume).

The Hellenistic World and Rome
For historians of the classical world, however, what is most distinctive
and significant about Daniel is the representation, through the medium of
narratives, dreams, and prophecies, of a succession of Near Eastern empires,
in steadily increasing detail and accuracy, from the Neo-Babylonian Empire through Achaemenid Persia to Alexander, and then to the Seleucids and
Ptolemies. Viewed literally, there are in these representations various mistakes, misidentifications, and transpositions. But, taken as a series, the picture presented by Daniel is indeed that which Momigliano argued should
be seen as a borrowing from the Greek world, namely the notion of history as a succession of world empires.31 In terms of the subsequent history
of the world the issues which were stirred up by Antiochus’ persecution and
the Maccabean counter-revolution, and which gave rise to the composition
of the book of Daniel as we have it, had a significance far greater than any
other aspects of the Hellenistic period. For it was in the persecution of the
s and the resistance to it that Jewish monotheism, its sacrificial cult, and
the personal observances required of its adherents faced and survived their
greatest test. But even as regards ancient political history, and the representation in literature of the successive regimes which had claimed domination
over the peoples of the Near East, Daniel can add a dimension, and the sense
of a longer perspective, which even Polybius was not in a position to attach
to those drastic swings of fortune which took place in  .., the year at
which his great History had originally been intended to end. But, as history,
Daniel’s brief and allusive representations of selected episodes, taken largely
out of context, and seen from a very precise and limited viewpoint, cannot
of course compare with Polybius’ profound conception of how events in different parts of the oikoumenē (inhabited world) had come to be interlinked,
and to form a single causative sequence. For an understanding of Hellenistic
history we will still depend on Polybius, and on his greatest modern interpreter.
. See A. D. Momigliano, ‘‘The Origins of Universal History,’’ Ann. Sc. N. Sup. Pisa,
ser. , . (): – On Pagans, Jews and Christians (), –.
 
The Background to the Maccabean Revolution:
Reflections on Martin Hengel’s
‘‘ Judaism and Hellenism’’ *
Introduction
The truism that important events are understood best when considered at
some distance in time may serve as an excuse for surveying only so belatedly the vast contribution to Jewish and Hellenistic history made by Hengel’s
major work, first published in German in , revised and enlarged in ,
and issued in an English translation in .1 If some alternative perspectives
are suggested in what follows, and some doubts expressed about the central
concept of a Hellenistic reform movement on the part of a group within the
Jewish community, that implies no lack of admiration or gratitude for what
is, along with P. M. Fraser’s Ptolemaic Alexandria (), the most important
contribution to the history of the Hellenistic world since M. I. Rostovtzeff ’s
Social and Economic History (). More specifically, as a study of the interaction of Judaism and Hellenism, it has been rivalled in recent years only by
the great work of Menahem Stern, Greek and Latin Authors on Jews and Judaism I–III (–),2 and by the penetrating essay of A. D. Momigliano in
chapters – of Alien Wisdom (). In looking back on Hengel, it will be
possible to take this and other subsequent work into account; but the im* First published in Journal of Jewish Studies  (): –. Versions of this paper were given
at seminars at the Institute of Classical Studies, University of London, and the Oxford Centre for Postgraduate Hebrew Studies. I am very grateful to Tessa Rajak and Geza Vermes
for helpful comments.
. M. Hengel, Judaism and Hellenism: Studies in Their Encounter in Palestine during the Early
Hellenistic Period I–II (), henceforth Hengel. Cf. also his Juden, Griechen und Barbaren:
Aspekte der Hellenisierung des Judentums in vorchristlicher Zeit ().
. See Tessa Rajak, ‘‘The Unknown God,’’ JJS  (): –.


The Hellenistic World and Rome
portant task is to try to sum up Hengel’s achievement, and to ask from what
different perspectives we might approach the events of the s—which on
any view were fundamental to the future of Judaism, and hence of Christianity and Islam. If there ever was an identifiable ‘‘turning point in history’’
it was the years  to  .. in Palestine.
Hengel’s Interpretation
It is the essential characteristic of Hengel’s approach that, in spite of the immense historical learning and real historical understanding employed in it,
his thesis is that of a Christian theologian: that the early Hellenistic period
saw a significant process of mutual assimilation and comprehension between
Judaism and paganism, which was brought to a halt by the nationalistic reaction under the Maccabees, and was only resumed and brought to fruition
in the preaching of Christianity to gentiles. To arrive at this conclusion Hengel starts (–) with a survey of Hellenistic political, administrative, economic, and cultural influences in Palestine which represents one of the best
studies of any region of the Hellenistic world. It would not be possible to
point to any general survey of a similar quality for Anatolia, Syria proper,
Babylonia, or Egypt. The conclusion (–) is that Palestinian Judaism was
as much a ‘‘Hellenistic’’ Judaism as that of the diaspora.
From there Hengel moves to Greek influence on the Jewish literature
and philosophy of the Hellenistic period, and to movements within Judaism which he takes as being in reaction to this, the growth of hasidism and
apocalyptic and the emergence of Essenism—which itself, in his view, had
Hellenistic features. The final section, on the Hellenistic re-interpretation of
Judaism and the reform attempt, begins, perhaps surprisingly, with literary
evidence for Greek views, starting with Hecataeus of Abdera, of the Jews as a
species of philosophers; it then moves to the (very slight) evidence for Jewish
identifications of the Jewish God with the supreme God (Zeus or Dis) worshipped by the pagans; indeed the evidence from the Hellenistic period is
confined precisely to two Alexandrian writers, Aristobulus and the author of
the Letter of Aristeas. Finally, Hengel comes to a sketch of the reform movement of – itself and an attempt, following E. Bickermann, Der Gott
der Makkabäer (), to identify the ideology of the reforms and the nature
of the cult briefly established in the Temple. It is the fact, which Hengel of
course clearly admits (), that we cannot securely establish the nature of
this cult which is the fundamental weakness of the book’s main thesis. For
we do not know, though we may wish to argue, that it was in any sense syn-
The Maccabean Revolution

cretistic. The evidence is perfectly compatible with the possibility that it was
quite simply a pagan cult.
We may postpone this question in order to look first at some other aspects of this complex series of events. Firstly, Hengel considers only briefly
(–) the parallels afforded by other Semitic deities and their identification with Greek gods. More significantly, he does not consider Hellenistic
influences on the Jews as part of the much wider, and in many respects obscure and mysterious, process of the fusion of cultures in the whole south
Syrian region. But it would only be if we could understand their immediate neighbours, the Phoenicians, Samaritans, Idumaeans, Nabataeans, and the
(presumably) Aramaic-speaking peoples of Transjordan and southern Syria
that we could begin to ask how distinctive the Jews were, and what if any
type of fusion with Hellenism could have been expected of them. Though
the book of course contains references to the cultures of these peoples, they
do not serve to provide a perspective for it. In the next section an attempt
will be made to sketch some possible contributions of this perspective.
An incomparably more glaring omission is that Hengel does not attempt
to portray the ordinary, day-to-day Judaism and its practises which the reforms set out to transform or abolish: ‘‘his book really deals with the Hellenization of an unknown entity.’’ 3 Momigliano’s phrase justly characterises
the book (and also Hengel’s Juden, Griechen und Barbaren), but not the state
of our knowledge. The subsequent section (‘‘The Jewish Community in the
Early Hellenistic Period’’) will sum up briefly some aspects of the structure
and nature of the Jewish community in the early Hellenistic period. Even a
slight sketch will show how un-Hellenistic it was and remained, in political
institutions, in observances, and in its culture and literary products.
One fact on which some suggestions will need to be made is the relation of the Ptolemaic and Seleucid kings to the high priesthood. Here too
it is a weakness of Hengel’s book that, again following Bickermann, he accepts with only cursory discussion (–) the unimportance of the role of
Antiochus IV Epiphanes. But even if (as is beyond question) the initiatives
taken by the Hellenisers were of fundamental importance, it remains the case
that the interruption of the Temple cult and the prohibition of Jewish observances were enforced by royal orders and carried out by royal employees.
. A. D. Momigliano, in his review in JThS  (): – Quinto Contributo alla
storia degli studi classicia del mondo antico (), –; for a further perspective, cf. his ‘‘J. G.
Droyser between Greeks and Jews,’’ History and Theory  (): – Essays in Ancient
and Modern Historiography (), –.

The Hellenistic World and Rome
We cannot avoid returning (‘‘Antiochus Epiphanes and Judaism,’’ below) to
the question of whether Antiochus had a ‘‘policy’’ of his own, and if so why
and to what end.
But the central question remains (last section but one): what was the nature of the cult or cults observed in the Temple between  and  ..? We
can understand (more or less) the nature of the faith which triumphed in the
Maccabean revolution. If we do not understand exactly what was the nature
of the cult which was supposed to replace it, it would be better to say so,
and not to build vast historical interpretations on hypotheses as to matters
of fact.
Greek and Native Cultures in the Syrian Region
As mentioned above, only new evidence could improve Hengel’s portrayal
of Hellenism in Judaea itself. One such item is indeed the bilingual ostracon
from Khirbet el-Kôm, between Hebron and Lachish, with brief Aramaic and
Greek texts acknowledging a loan.4 This third-century .. text (probably of
 ..) is notable as the earliest Greek document from Palestine, as attesting the loan-word QPYLS, from the Greek kapelos, a trader, and as a perfect
exemplification of the commercial influences which Hengel emphasises.
But for a fuller understanding of the phenomena of Hellenism in Palestine we need a wider perspective. Firstly, it is noticeable that on the fringes
of the Syrian region, in Nabataea and Palmyra, flourishing mixed cultures
could grow up in the late Hellenistic and early Roman periods, employing
both Greek and dialects of Aramaic in their public documents, exhibiting
exotic and magnificent variants of Greek architecture, and observing cults in
which Greek elements intermingled with indigenous ones.5 In Syria proper,
notably at Hierapolis/Bambyce and at Heliopolis/Baalbek, it is notorious
that there survived within Greek cities temples and cults which both were,
and were perceived at the time to be, entirely non-Greek in origin and character. The obvious instance is Lucian’s description in On the Syrian Goddess
of the cult of Atargatis/Hera at Hierapolis, with references to that of Melqart/Hercules at Tyre and to cults at Heliopolis (which he thought to be of
Egyptian origin), Byblos, and Apheca. Similarly, when Herodian comes to
. L. T. Geraty, ‘‘The Khirbet el-Kôm Bilingual Ostracon,’’ BASOR  (): –;
cf. A. Skaist, ‘‘A Note on the Bilingual Ostracon from Khirbet el-Kom,’’ IEJ  (): –.
. For Nabataea, see, e.g., the brief survey and bibliography in Schürer, Vermes, and
Millar, History I, app. ; and P. C. Hammond, The Nabataeans: Their History, Culture and
Archaeology (). For Palmyra see M. A. R. Colledge, The Art of Palmyra ().
The Maccabean Revolution

describe the coup d’état which brought to the imperial power in .. 
a great-nephew of Septimius Severus through his Syrian wife Julia Domna,
he gives a perfect description of a non-Greek cult as preserved at Emesa and
integrated in the life of Greek Syria:
Both boys (the pretender and his cousin) were dedicated to the service
of the sun god whom the local inhabitants worship under its Phoenician name of Elagabalus. There was a huge temple built there, richly
ornamented with gold and silver and valuable stones. The cult extended
not just to the local inhabitants either. Satraps of all the adjacent territories and barbarian princes tried to outdo each other in sending costly
dedications to the god every year. There was no actual man-made statue
of the god, the sort Greeks and Romans put up; but there was an enormous
stone, rounded at the base and coming to a point at the top, conical
in shape and black. This stone is worshipped as if it were sent from
heaven; on it there are some small projecting pieces and markings that
are pointed out, which the people would like to believe are a rough
picture of the sun, because this is how they see them. Bassianus, the
elder of the two boys, was a priest of this god. . . . He used to appear in
public in barbarian clothes, wearing a long-sleeved ‘‘chiton’’ that hung
to his feet and was gold and purple. His legs from the waist down to
the tips of his toes were completely covered similarly with garments
ornamented with gold and purple. On his head he wore a crown of
precious stones glowing with different colours.6
We could not ask for a better model of how the high priest of an aniconic cult
in the city of Jerusalem might have appeared to a Greek observer if the hypothetical transformation had taken place, and the cult preserved in its ancestral
form but integrated into that of a Supreme God compatible with the pantheon of Graeco-Roman paganism. The fact that (rightly or wrongly) Herodian perceived the cult in the inland city of Emesa as ‘‘Phoenician’’ is also
significant, for two main reasons. Firstly, the Phoenician cities are the clearest
examples of pre-Greek cities which retained their traditions and identities
while also becoming integrated in the Hellenistic and Roman world.7 Phoenician inscriptions continue until the reign of Augustus, Phoenician lettering on the coins of Tyre until towards the end of the second century ..
. Herodian , , –, trans. C. R. Whittaker in the Loeb ed. For the cult, see RAC, s.v.
‘‘Elagabal.’’ For Emesa, H. Seyrig, ‘‘Caractères de l’histoire d’Emèse,’’ Syria  (): –.
. See ‘‘The Phoenician Cities: A Case-Study of Hellenisation,’’ Proc. Camb. Philol. Soc.
 (): – ( chapter  in the present volume).

The Hellenistic World and Rome
Without our entering here on the complex problems involved, it is enough
to say that Philon of Byblos at the beginning of the second century ..
claimed to have translated the work of an ancient Phoenician, Sanchuniathon, on the Phoenician cults, and that his translation was used by Porphyry
of Tyre in the third century and then by Eusebius of Caesarea. Not only,
therefore, was there a continuity of cults, with much assimilation to Greek
deities, but also (however slight) a continuous literary tradition.8 Moreover
Eusebius, perhaps quoting Porphyry, can confirm the accuracy of the literary
tradition by reference to still-existing cult practices: the testimony of the
ancient theologi drew confirmation ‘‘from the nomenclature of the gods still
applied to this day in the cities and villages in Phoenicia and the interpretation given to the mysteries celebrated among each people.’’ 9
Secondly, not only did these now Graeco-Phoenician cities continue to
exercise, as they always had, a cultural influence on Judaea (most aptly symbolised by the exclusive use of Tyrian shekels for the Temple dues), but communities inland in the immediate neighbourhood of Judaea tended to identify themselves as Phoenician, or rather ‘‘Sidonian.’’ The best documented
case, well discussed by Hengel (, ) is that of the community which in
the second century .. described themselves, in Greek, as ‘‘the Sidonians
in Marisa,’’ in Idumaea. The second instance presents crucial difficulties, all
the more so because, as we shall see, it is the cornerstone of the entire interpretation of the ‘‘reform movement’’ as understood by Bickermann and
Hengel.
 Maccabees :– tells us that Antiochus sent an emissary to profane the
Temple of Jerusalem and dedicate it to Zeus Olympius, and that of Mount
Gerizim to Zeus Xenius, ‘‘as those who dwelt in the place requested.’’ Josephus, who did not use  Maccabees, appears to present us with the petition
itself,10 in which they claim to be Sidonians, unrelated to the Jews, and ask for
their temple to be known as that of Zeus Hellenius; and also with the king’s
instruction to his agent, referring to them as ‘‘the Sidonians in Shechem.’’
All would be well if we could, with Momigliano, be confident that this and
the book of Daniel represent the ‘‘only two certain pieces of contemporary
evidence for the religious situation of the persecution period.’’ 11 That may
indeed be so. But we have to reckon with the fact that the only source for
. The evidence, which demands further study, is summarily collected in F. Millar, ‘‘Paul
of Samosata, Zenobia and Aurelian: The Church, Local Culture and Political Allegiance in
Third-Century Syria,’’ JRS  (): –, on – ( chapter  of the present volume).
. Eusebius, Praep. Ev. , ,  F. Jacoby, FGrH  F.  ().
. Ant. , –.
. Alien Wisdom, .
The Maccabean Revolution

the document is not itself contemporary. Furthermore, Josephus quotes it
explicitly as evidence for a thesis which he has put forward earlier,12 that the
Samaritans claimed or denied relationship to the Jews according to the advantages of the moment. Thirdly, there is no other evidence recording either
a reversion of the Samaritan Temple to its original cult after the end of persecution or its continuance as a temple of Zeus Xenius or Hellenius until its
destruction by John Hyrcanus.
A substantial doubt must therefore remain. But  Maccabees at least attests the fact of a petition for the transformation of the cult on Mount Gerizim; and we have the independent evidence of Pseudo-Eupolemus, perhaps
writing in the second century .., who explains ‘‘Argarizin’’ as ‘‘mountain
of (the) Highest’’ (hypsistos):13 ‘‘Zeus Hypsistos,’’ being, as is well known, a
possible way of referring to the Jewish God in the Graeco-Roman period.
We still cannot be certain of course what Pseudo-Eupolemus himself implied by this, or whether his view was shared by those who worshipped there.
However, we have a considerable body of evidence, admittedly from varying regions of Syria and varying dates, which clearly shows that indigenous
cults could be preserved and integrated with their now Hellenised environment without losing their identity or continuity. What is more, a study of
the Semitic cults of the Syrian region, and in particular of private dedications, whether in Greek or in dialects of Aramaic, argues that the evidence
reflects a growth of the conception of a single supreme god, addressed in
various names.14 If this evidence will require further consideration, it is still
clear that, if the cult in Jerusalem were sufficiently similar to the native cults
persisting elsewhere in the Syrian region, it might have undergone a comparable assimilation provided that the social and cultural conditions for this
were also present. Whether they were present is a question which can only be
answered by looking at the structure of the Jewish community in the period
preceding the ‘‘reform’’ and revolution.
The Jewish Community in the Early Hellenistic Period
The Persian period, and the reforms of Ezra and Nehemiah, had left the
Jews as a small tightly knit community centred on the Temple, occupying a
. Ant. , .
. Eusebius, Praep. Ev. ,  Jacoby, FGrH . See also Hengel, –, and A.-M.
Denis, ‘‘L’historien anonyme d’Eusèbe (Praep. Ev. , –) et la crise des Macchabées,’’ JSJ
 (): –. I am unable to discern why the mention of Mount Gerizim should lead
to the presumption that the author was a Samaritan.
. J. Teixidor, The Pagan God: Popular Religion in the Greco-Roman Near East ().

The Hellenistic World and Rome
restricted area round Jerusalem and in uneasy relations with their Samaritan neighbours.15 The descendants of the House of David disappear in this
period, leaving the political direction of the community in the hands of the
high priests, the succession of whom can now, thanks to the Samaria papyri,
be more confidently reconstructed.16 In addition to the high priest there was
also a local Persian governor ( peḥah).
No historically valid account survives of the absorption of the Jewish
community into the empire of Alexander and his successors. In default of
any other satisfactory context, therefore, the well-known series of Yehud
coins, one of them bearing the name ‘‘Yeḥezk iyo,’’ has been considered as belonging to the late Persian period, as has been one with a longer inscription,
‘‘Yeḥezk iyo ha-peḥah.’’ 17 But now there has been published a series of coins
with the legend ‘‘Yehudah,’’ in a script similar to the latter and showing the
portrait of Ptolemy I.18 Since Ptolemy did not mint coins with his portrait
until  .., the coin must relate to the period after  when he finally
regained control of Palestine. The coins therefore tend to suggest a continuity of Judaea as a political unit into the Ptolemaic period. More than that,
however, as Kindler saw, they raise once again the question of the ‘‘Hezekiah,
High Priest of the Jews,’’ whom Hecataeus of Abdera mentions as in office at
the time of Ptolemy’s earlier possession of Palestine, after the battle of Gaza
in  ..19 According to Hecataeus, this man was among a number of Jews
who, impressed by Ptolemy’s mildness and moderation, accompanied him
to Egypt after the battle (the Greek does not imply that he emigrated permanently to Egypt). More important, Josephus quotes from Hecataeus the
following sentence: ‘‘This man, having gained this honour and become our
companion, summoning some of those with him read to them the whole
scroll, which contained a written description of their settlement and constitution.’’ 20
What was the ‘‘honour’’ (timē ) which Hezekiah had received (from
. For a clear treatment, see J. Bright, History of Israel 2 (), chaps. –. Note also
the interesting, if very speculative, treatment by Morton Smith, ‘‘Palestinian Judaism in the
Persian Period,’’ in H. Bengtson, The Greeks and the Persians (), chap. .
. See F. M. Cross, ‘‘A Reconstruction of the Judean Restoration,’’ JBL  (): –.
. L. Y. Rahmani, ‘‘Silver Coins of the Fourth Century .. from Tel Gamma,’’ IEJ 
(): –.
. A. Kindler, ‘‘Silver Coins Bearing the Name of Judea from the Early Hellenistic
Period,’’ IEJ  (): –; D. Jeselsohn, ‘‘A New Coin Type with Hebrew Inscription,’’
IEJ  (): –.
. Quoted by Josephus, C. Ap. , – Stern, Greek and Latin Authors I, no. .
. C. Ap. , , reading διφθέραν with Stern (n. ).
The Maccabean Revolution

Ptolemy?)? We must remember that at some point after Alexander’s conquest
the governor ( peḥah) of the Persian period disappears, leaving the high priest
as the political head of the community. The corollary of that, however, was
that in the Seleucid period (from  ..), or at least by the beginning of Antiochus Epiphanes’ reign (from  ..), the high priest was appointed by the
king.21 Is it possible therefore that the ‘‘honour’’ was indeed the high priesthood, and that it had been taken into the king’s gift as early as the beginning
of the Ptolemaic period, in compensation for the abandonment of the appointment of a separate governor? That would readily explain why different
sources refer to this Hezekiah as high priest and as governor; and the dual
function may be reflected in what Joseph the Tobiad is reputed as saying to
Onias in a scene which seems clearly to belong to the Ptolemaic period, more
precisely to the reign of Ptolemy III Euergetes (– ..), though misplaced in Josephus’ narrative. Joseph reproaches Onias (on this view Onias II)
for withholding tribute due to the king, ‘‘on account of which,’’ Joseph said,
‘‘he had received the chief magistracy [ prostasia] and had obtained the high
priestly office.’’ 22 It would not have been impossible that peḥah could have
remained in use to denote the secular functions now assumed by the high
priest. Ptolemy himself, before his assumption of the royal title, is referred
to by the Persian title ‘‘satrap’’ in a Greek document of  .. from Egypt
(P. Eleph. ).23
If these were the circumstances, then it would not have been surprising if
there had not been unanimity as to who had been the legitimate high priest
at any one time, and hence that Hezekiah did not achieve a mention in the
list of high priests in Josephus. Hecataeus’ evidence is further valuable for
providing the view-point on the Jewish community of a sympathetic and
curious gentile observer at the beginning of the Hellenistic period. He emphasises above all the rigorous observance of the Law, the privileges of the
priests (, in number), the hostility to pagan cults, and the centrality and
importance of Jerusalem and the Temple. In another passage, transmitted by
Diodorus, he emphasises the absence of a cult image, the leading role of the
priests, the absence of a king (believing that there had never been one), and
the importance of the high priest.24 If Hecataeus also imported a number of
idealising Greek notions into his portrait of Jewish society, it is none the less
important to stress the extent to which he saw it as a community whose struc. Josephus, Ant. , –.
. See Ant. , , Loeb trans. See Schürer, Vermes, and Millar, History I, .
. Sel. Pap. I, no. .
. Diodorus ,  Stern, Greek and Latin Authors I, no. .

The Hellenistic World and Rome
ture and conventions were fundamentally non-Greek. We lack a similarly
informed witness from the end of the third century (though Agatharchides
in the second, looking back to Ptolemy I’s conquest of Judaea, gives a quite
detailed description of the observance of the Sabbath).25 But in any case the
distinctive features of Jewish society are clearly documented in the ‘‘Seleucid
charter of Jerusalem,’’ Antiochus III’s letter to an official called Ptolemaios,
laying down Jewish privileges, granted as a reward for their favourable reception of him when he took Palestine from the Ptolemies in  .. After
listing provisions for sacrifice and materials for building work on the Temple,
he continues, ‘‘And all the members of the nation shall have a form of government in accordance with the laws of their country, and the senate [ gerousia],
the priests, the scribes of the temple and the temple singers [Levites] shall
be relieved of the poll tax.’’ A subsequent proclamation gave royal backing
to the prohibition on entry to the Temple for gentiles and unclean Jews, and
forbade the import into Jerusalem of the meat or skins of animals declared
unclean by the dietary laws.26
The work of Ben Sira, written some time in the high priesthood of
Simon II (late third or early second century ..), is notable as an original
work composed in Hebrew, as emphasising two distinctive features of contemporary Jewish society, the class of hereditary priests and the scribes, and as
reflecting (esp. chaps. –) a deep sense of continuity with the past.27 Very
similar characteristics mark the book of Daniel, which indisputably reached
its final form in the mid-s. The documents from Qumran show that these
were not isolated features. Without going into details beyond the competence of the author, it will be enough to emphasise that the Qumran texts
include parts of every old Testament book except Esther, and a substantial
range of original works related to biblical themes and written in this period
is in either Hebrew or Aramaic.28 For instance, no reader of the fascinating
fragments edited by J. T. Milik 29 will be readily disposed to assent without
severe qualification to the proposition that Palestinian Judaism was as Hel. Josephus, C. Ap , – Stern, Greek and Latin Authors I, no. a.
. Preserved only by Josephus, Ant. , –. But note the documents relating to this
Ptolemaios in the inscription published by Y. H. Landau, ‘‘A Greek Inscription Found near
Hefzibah,’’ IEJ  (): – ( SEG XXIX ).
. See E. Rivkin, ‘‘Ben Sira—The Bridge between the Aaronide and Pharisaic Revolutions,’’ Eretz-Israel  (): *–*. Part of the original Hebrew text has, of course,
been recovered at Masada; see Y. Yadin, ‘‘The Ben Sira Scroll from Masada,’’ in Masada: Final
Reports VI (), –.
. See up-to-date surveys in DJD XXXIX ().
. J. T. Milik, The Books of Enoch: Aramaic Fragments of Qumrân Cave  ().
The Maccabean Revolution

lenistic as that of the diaspora. On the contrary, what we should emphasise
is the uniqueness of the phenomenon of an original and varied non-Greek
literary activity developing in a small area only a few miles from the Mediterranean coast. The preservation of a native culture in Egypt or in distant
Babylonia is by no means so surprising a phenomenon.
What is more, it is precisely the nature of the first phase of the Hellenising
movement after  .., instituted beyond question by elements within the
Jewish community, which shows how un-Greek Jerusalem had remained up
to that moment. It is important to emphasise the earliest phase of the Hellenising movement—that which preceded the robbery of the Temple treasures
and the forcible persecution of Judaism—because it is here that we have the
clearest evidence of initiatives from within the community; because it shows
the relation of the Seleucid king to the high priesthood; and precisely because it shows the novelty of Hellenistic ways of life in the Jerusalem of the
s.  Macc. :– (followed by Josephus, Ant. , –) preserves only
the overall ideology of ‘‘making a covenant with the peoples round about,’’
and the details of the establishment of a gymnasium and concealment of circumcision. For a fuller conception we have to turn to  Maccabees, the subject of an excellent German translation and commentary by Chr. Habicht
(the first time, it should be noted, that this major product of Hellenistic historiography has been studied in detail by an expert in Hellenistic history
and epigraphy).30 It is worth looking again at the much-quoted passage of
 Macc. :– which describes the events following the death of Seleucus IV
Philopator in  ..:
But when Seleucus departed this life and the kingdom was taken by
Antiochus, called Epiphanes, Jason the brother of Onias, usurped the
high priesthood by illegitimate means, promising the king in a petition
 talents of silver, and  talents of other revenue. He undertook beyond this to pay a further  talents if he were granted permission to
establish by his own authority a gymnasium and a corps of ephebes and
to enrol those in Jerusalem as ‘‘Antiochenes.’’ 31 When the king agreed,
. Jüdische Schriften aus hellenistisch-römischer Zeit I: Historische und legendarische Erzählungen ; Chr. Habicht, . Makkabäerbuch ().
. The translation of this phrase is a notorious crux. I take it to mean that all the inhabitants of Jerusalem were to become citizens of a new polis called (like many others)
‘‘Antiochia.’’ For the clearest statement of this view, and a refutation of Bickermann’s interpretation, that the reference is to the establishment of a corporation of ‘‘Antiochenes’’
within the population of Jerusalem, see G. Le Rider, Suse sous les Séleucides et les Parthes (Mém.
Miss. Arch. Iran. , ), –.

The Hellenistic World and Rome
and he gained the office, he immediately set about converting his fellow countrymen to the Greek way of life. Abolishing the existing royal
privileges gained by Iohannes, the father of the Eupolemus who (later)
undertook an embassy to Rome for friendship and alliance—and the
legitimate institutions, he brought in illegal customs. For he saw fit
to establish a gymnasium below the acropolis and lead there the most
athletic of the ephebes wearing sunhats. There was such a flowering of
Hellenism and advance of gentile customs through the overwhelming
wickedness of the impious Jason, no true high priest, that the priests
were no longer conscientious over the duties concerned with the sacrifice, but, despising the Temple and neglecting the sacrifices, they hastened to take part in the unlawful exercises in the palaestra as soon as
the sound of the discus summoned them.
It is noticeable in this account that the events depended on the right of the
king to dispose of the high priesthood; that the services of the Temple continued, if neglected to some degree; that there had been no previous gymnasium in Jerusalem; and that even the wearing of the Greek sunhat ( petasos)
was regarded as an outrageous novelty. The testimony of  Maccabees and
Josephus (above) would also confirm that the Hellenisers and their followers
had all been circumcised, a step which they attempted to reverse.
None the less, there is no sign at this stage of any change in Temple ritual,
of any enforcement by Seleucid officials or of any popular resistance. These
further steps came when the possibility of acquiring the high priesthood
from Antiochus had led to the replacement of Jason by Menelaus ( Macc. :
–). It is important to emphasise that this is a new phase. There is no evidence to connect Menelaus with the Hellenising movement described above,
except that his opportunity to gain office from Antiochus came when he was
sent to take money to him from Jason (:). In his time as high priest, if we
may follow the only detailed narrative ( Macc. :–:), there was oppression in Jerusalem by him, civil war between his party and that of the deposed Jason, culminating during Antiochus IV Epiphanes’ invasion of Egypt
in  ..,32 and finally (/) the entry of the king and his forces into Jerusalem, a massacre of the inhabitants, entry into the Holy of Holies by the
king, and robbery of the Temple treasures. When the king departed, he left
an overseer (epistates) in Jerusalem and another over the Samaritans at Mount
Gerizim ( Macc. :–).
. The chronology is, of course, much disputed. I would however retain that of
Schürer, Vermes, and Millar, History I, –.
The Maccabean Revolution

Drastic and violent changes thus intervened between the Hellenising
movement initiated by Jason, and the desecration of the Temple and forcible
persecution of Judaism which followed under his rival and successor, Menelaus, in – .. In between came Antiochus’ invasion of Egypt in ,
stopped outside Alexandria by the famous Roman mission led by Popillius
Laenas. It may not be irrelevant in the context to note that truly remarkable evidence, in the form of demotic dream texts from Egypt, supports the
testimony of Porphyry and Jerome on Daniel, and of a Greek papyrus, that
in this phase Antiochus was claiming the kingship of Egypt, was crowned at
Memphis, and acted as king in Upper Egypt.33
For the author of Daniel (:) clearly saw some connection between the
rebuff of Antiochus by the ‘‘ships of Kittim’’ and the persecution which followed. What Antiochus’ motives may have been will be discussed in the next
section, and what cult was established in the Temple in the last one. What
is relevant at this stage is to emphasise, firstly, that the Temple cult had continued apparently unaffected through the phase of Hellenism and the depredations of Antiochus in /; and secondly, and more important, the extent
of the Jewish observances which the persecution attempted to eradicate. To
say this is not to deny the truth of Hengel’s demonstration (–) that
many Jews apostatised, went over to the Greek side, and aided the persecution; as Hengel shows, it was largely at them that the Maccabean movement
was aimed. None the less, the persecuting measures were directed against the
keeping of the Sabbath and festivals, the offering of the traditional sacrifices
in the Temple, circumcision, and possession of the Torah ( Macc. :–;
 Macc. :–; Josephus, Ant. , –). The narratives also presuppose
observance of the dietary laws, at least in respect of pork.
Whether any Jews played an active role in initiating this phase of the conflict (as opposed to the Hellenising movement under Jason) is a question to
which we shall have to return. What is clear on the one hand is that the persecution involved something much wider than the transformation or substitution of a cult in the Temple; the Maccabean resistance arose in response to
attempts to change by force the private observances of the populace. It is precisely the evidence for these observances, and of the resistance to abandoning
them which, along with that previously mentioned, indicating the continuity of the Jewish community from the post-Exilic period, shows how misleading it is to take some Hellenistic influences on the literary or philosophical products of Judaism and deduce from them that Judaism itself had become
‘‘Hellenistic.’’ Though it is possible to find parallels, in Syria and Egypt, for
. See J. D. Ray, The Archive of Hor (), esp. –.

The Hellenistic World and Rome
circumcision 34 and the avoidance of pork,35 the existence of a complex set
of observances binding (in principle) on the whole population has no parallel; nor do we yet know of any other Near Eastern people speaking a Semitic
language who in the Hellenistic period generated a whole range of works in
different genres in their own language (or languages, Hebrew and Aramaic).
Finally, if slight traces reveal (text to n.  above) that the Phoenicians still
possessed a historical tradition of their own, nothing parallels the existence
of a sacred book which was at the same time a national history, and which,
as Ben Sira, Maccabees, and the Qumran documents all show, was in active
circulation among the people and was the primary agent in forming their
consciousness.
The proclamation of Antiochus III had explicitly recognised both the distinctive socio-political status and the distinctive observances of the Jewish
people. Have we enough evidence to say whether the events under Antiochus IV were the product of a new policy or conviction, and if so why and
to what end?
Antiochus Epiphanes and Judaism
It is a striking feature of modern historiography that there should be so
strong a tendency to look to sources other than the reigning Seleucid king
himself for the explanation of a persecution carried out by royal command
and by royal agents.36 The presumption that changes should be seen in terms
of the wider structures and ideologies of society is not easily reconciled with
the fact of monarchy. Yet both pagan and Jewish sources reflect the view
that Antiochus did intend some overall change in Jewish observance. Diodorus, in retailing the siege of Jerusalem by Antiochus VII Sidetes, used a story
which represented the king’s advisers as reminding him of how Epiphanes
had entered the Holy of Holies and found an image which he supposed to
be that of Moses, who had created the misanthropic customs of the Jews:
And since Epiphanes was shocked by such hatred directed against all
mankind, he had set himself to break down their traditional practices.
Accordingly, he sacrificed before the image of the founder and the
open-air altar of the god a great sow, and poured its blood over them.
Then, having prepared its flesh, he ordered that their holy books, con. See the references in Schürer, Vermes, and Millar, History I, –.
. See, e.g., Bickermann, Der Gott der Makkabäer, .
. So, before Hengel, Bickermann, Der Gott der Makkabäer, ff.; but cf. V. Tcherikover,
Hellenistic Civilisation and the Jews (), ff.
The Maccabean Revolution

taining the xenophobic laws, should be sprinkled with the broth of the
meat; that the lamp, which they call undying and which burns continually in the temple, should be extinguished; and that the high priest
and the rest of the Jews should be compelled to partake of the meat.37
Sidetes, however, was not moved but made peace with the Jews. What concerns us, however, is that the story, for all its mythical elements, embodies the
conception, followed later by Tacitus, Hist. , , on the part of a pagan author,
that Antiochus had been seized by the conviction that Jewish observances
were undesirable and should be abolished. It should be noted, however, that
there is no hint from this pagan source of ‘‘syncretism.’’ What was involved
in the view of Diodorus was abolition.
Now it is a notorious (and true) commonplace that ancient states, though
they controlled (and might therefore reject, or attempt to reject) the importation of foreign cults, fundamentally respected the existing gods and cults of
different localities. The Ptolemies for instance, regulated and controlled the
revenues of Egyptian temples. But there was no question of their attempting
to abolish them and institute Greek gods and temples. In Syria, as we saw
earlier, non-Greek temples and cults survived the Seleucid period and were
famous and frequented under the Roman Empire. In Babylonia cuneiform
documents associated with the major temples persist through the Seleucid
period; a particularly relevant one shows Antiochus I Soter (–/ ..)
proclaiming his restoration of the temples of Esagila and Ezida.38 The only
hint of a different attitude comes in an illustration given by Pausanias of the
justice and piety of Seleucus I (– ..); ‘‘when he founded Seleucia on
the river Tigris, and brought Babylonian colonists to it, he left standing both
the walls of Babylon and the sanctuary of Bel, and allowed the Chaldaeans
to dwell round the sanctuary as before.’’ 39 The passage contains just a suggestion of presumptions which might have been relevant when a city called
‘‘Antiochia’’ was established in place of Jerusalem. But in fact Seleucus had
permitted the continuation of the local cult; and in Jerusalem not only had
Antiochus III given privileges which specifically reflected the Temple cult,
and the structure of the society related to it, but (according to  Macc. :)
Seleucus IV Philopator (– ..) had paid for the sacrifices out of his
own funds.
What happened under Antiochus Epiphanes was therefore both an exception to an undeniable general pattern and a break with the established
. Diodorus /, I (Loeb trans.) Stern, Greek and Latin Authors I, no. .
. J. B. Pritchard, Ancient Near Eastern Texts (), .
. Pausanias, Description of Greece , , , trans. Frazer.

The Hellenistic World and Rome
relations between the Seleucid kings and Jerusalem. Nor is there any obvious
explanation for it in the patterns of Antiochus’ behaviour towards the wider
Greek world: his generosities to cities and temples show nothing more than
the display which was essential to the role-playing of any ambitious Hellenistic monarch; and the evidence for his foundation of new Greek cities
is less than has been supposed.40 But two innovations on the part of Antiochus are undeniable: he was the first Seleucid king to use the title ‘‘god
manifest’’ (Theos Epiphanēs) documented on coins and an inscription; and he
moved away from his dynasty’s traditional devotion to Apollo to a particular reverence for Olympian Zeus, whose image replaced that of Apollo on
the coins of Antioch. He also built, or contributed to, temples of Olympian
Zeus elsewhere, as in Athens and Priene. Both phenomena are connected by
Bunge with his difficulties in establishing his regime after his usurpation in
.41 The notion that Antiochus identified himself as Zeus is now universally
rejected. But it is significant that the mint of Antioch, followed by many
others, portrayed him with a radiate head which was the recognised symbol
of the sun god (Helios).42 As always, it is a matter of speculation what effect
if any such changes in the iconography of a ruler had on the attitudes of his
subjects. Daniel, however, does allude, in obscure terms (:–), to some
changes in religious observance by Antiochus.
We have already noted the role of the king in the first phase of ‘‘Hellenism’’
in Jerusalem, which required no more on his part than the (as I suggest) established right of the appointment of the high priest by the king, a pressing need
for money, and that general attachment to the notion of Greek city life which
we would expect any Hellenistic king to have shared. Nor do the massacre
in Jerusalem and the robbery of the Temple treasures in / .., between
the Egyptian campaigns, need any very special explanation; Josephus, C. Ap.
, , can quote a number of pagan historians for the view that he did so
simply to obtain much-needed funds. What does require such an explanation
is the positive measures of  .. Our only contemporary source, Daniel,
sees these steps as a reaction to the rebuff by Popillius Laenas in  (:):
‘‘The ships of Kittim will set out and go against him, and he will be humbled.
And he will turn back and rage against the holy covenant. And he will take
action and turn and pay heed to those who desert the holy covenant.’’ Daniel
. See the valuable discussion by O. Mørkholm, Antiochus IV of Syria (), chaps. 
and .
. J. G. Bunge, ‘‘ ‘Theos Epiphanes,’ Zu den ersten fünf Regierungsjahren Antiochos IV
Epiphanes,’’ Historia  (): –.
. J. G. Bunge, ‘‘ ‘Antiochus-Helios,’ Methoden und Ergebnisse der Reichspolitik Antiochos’ IV Epiphanes von Syrien im Spiegel seiner Münzen,’’ Historia  (): –.
The Maccabean Revolution

thus sees the event as springing from the state of mind of the king, alluding
in passing, however, to his paying attention to ‘‘those who desert the holy
covenant.’’ He clearly alludes to the establishment of a gentile garrison in
Jerusalem, less clearly to a general prohibition of Jewish observances.  Macc.
:– also sees the events as initiated by the king, who began by sending
‘‘to the cities of Juda’’ an official with an army, and establishing a garrison.
The elegy which concludes the first section (:–) implies that already
at this stage the effect was a flight of the Jewish inhabitants of Jerusalem
and neglect of the Temple. Then there follows the statement (:) that the
king ‘‘wrote to all his kingdom that all should be one people and that each
(people) should abandon their own customs. All the peoples conformed to
the word of the king.’’ We have to admit that no other evidence survives for
any such general measure, or indeed any comparable measure affecting anyone except Jews and Samaritans. The description of the persecution which
follows (:–) shows both a prohibition of Jewish observances and of the
Temple cult, as mentioned above, and a positive enforcement of pagan observances, both in Jerusalem and elsewhere. It was when the royal officials
came to Modein to compel the people to sacrifice that Mattathias began the
Maccabean movement ( Macc. :–).
 Maccabees, always more circumstantial, describes three separate missions by royal officials to Judaea: a Phrygian named Philippus left as epistatēs (commander) in / (:), and one Andronicus as epistatēs over the
Samaritans (:); then the ‘‘mysarch’’ (an unintelligible term) Apollonius,
with a force which carried out a massacre (:–); and finally an Athenian named Geron ‘‘to force the Jews to give up their ancestral beliefs and no
longer order their lives according to God’s commands. And also to desecrate
the Temple in Jerusalem and dedicate it to Zeus Olympius, and that on Mount
Gerizim, as the inhabitants had petitioned, to Zeus Xenius.’’ 43  Maccabees
goes on to describe the positive and negative features of the persecution in
comparable terms, adding the vital detail that a decree was sent out to the
neighbouring Greek cities to enforce the same measures on the Jews there
(:–).
It is not necessary to consider in the same detail the secondary account
in Josephus (Ant. , –) except to note that it represents the Samaritan petition (text to n.  above) as a reaction to their exposure to the same
persecution. They address the king as ‘‘Basileus Antiochus Theos Epiphanes’’;
they assert that they are Sidonians, not Jews; claim that their observance of
some Jewish customs is adventitious (which the king’s memorandum takes
. For this reading of the phrase, see Habicht, . Makkabäerbuch, on this verse.

The Hellenistic World and Rome
as meaning that they will live according to Greek customs); and ask for their
temple ‘‘without a name’’ on Mount Gerizim to be dedicated to Zeus Hellenius.
Even if we take this document as essentially authentic (some doubts are
expressed in text following n.  above), it and the parallel reference in
 Maccabees remain the only evidence that anybody in the locality actually
requested that the temple should be dedicated to Zeus. No such evidence
exists for Jerusalem.
Thus our evidence, both pagan and Jewish, shows that what was involved
in the s involved (a) measures taken by officials and troops sent by Antiochus; (b) a renunciation of Jewish customs and observances by all those
regarded as following them, in Jerusalem, in the ‘‘cities’’ of Judaea (like
Modein), in the Samaritan territory, and (according to  Maccabees) in the
surrounding Greek cities (but not, so far as we know, among the Jewish communities of Greek cities elsewhere, such as Antioch); (c) an imposition on
individuals of acts of pagan worship; (d) an attribution to Zeus, under different names, of the temples on Mount Gerizim and in Jerusalem; and (e) (attested only for Jerusalem) a transformation of the cult in the Temple, of a
nature and purpose which have yet to be discussed.
The evidence of actual events would thus support the view of Diodorus,
or his source, that at any rate one motive which was involved was an intention on the part of Antiochus to abolish by force all observance of the Law.
This conclusion is strongly supported if we accept as authentic the royal letters collected out of chronological order in  Macc. .44 In spring  ..,
before the recapture of the Temple, Antiochus writes from his eastern expedition, to the ‘‘gerousia [council or council of elders?] of the Jews and the
other Jews’’ promising an amnesty to those who return home before a certain date and saying ‘‘the Jews may use their own way of life and customs as
before’’ (:–). Even more clearly, after Antiochus’ death near the end
of , his successor, Antiochus V Eupator, writes to an official: ‘‘Now that
our father has departed to the gods . . . learning that the Jews, not consenting to the adoption of Greek customs wished by our father, but preferring
their own way of life, demand that they should be granted the observance
of their laws, we, desiring that this people should be at peace, have decided
that their temple should be restored to them and that this community should
live according to the customs of their forebears’’ (:–).
By this time the Temple had already been re-captured and re-dedicated,
. For the most authoritative discussion, see Habicht, . Makkabäerbuch, –. Note
Habicht, ‘‘The Royal Letters in Maccabees II,’’ Harv. Stud. Class. Phil.  (): –.
The Maccabean Revolution

though a Seleucid garrison remained in Jerusalem; so Antiochus was in this
respect restoring what he no longer held. But the vital implication of both
letters is that there had indeed been a policy of conversion on the part of Antiochus, and that it concerned essentially the question of Jewish or of Greek
customs and observances.
The fact is, in my view, established. It is best to confess, however, that
there seems no way of reaching an understanding of how Antiochus came to
take a step so profoundly at variance with the normal assumptions of government in his time. We may note, however, that he did have a later imitator:
for it can hardly be an accident that Hadrian, the most phil-Hellenic (and
well read) of Roman emperors, was to forbid circumcision and to re-found
Jerusalem as a Roman colonia named after Jupiter Capitolinus.
There remain the two related questions: what evidence have we for the
influence of Jewish Hellenisers on Antiochus’ decision and for the nature of
the cult established in the Temple in – ..?
The Hellenisers and the Temple Cult of – ..
As we saw above, Menelaus had gained the high priesthood by outbidding
his predecessor, Jason, the leader of the earlier Hellenising movement. Nothing in our sources reveals his role at that stage, or (in relation to the question of Hellenism) during his high priesthood except that in the course of
a confused passage (Ant. , –) Josephus says that when pressed by the
deposed Jason (see above) Menelaus and the Tobiads—never mentioned in
this context by either book of Maccabees—went to Antiochus and said that
they wished to adopt Greek customs. But the whole context is muddled, and
we do not reach good evidence until Antiochus IV’s letter of spring  ..
(above), which shows Menelaus with the king on his campaign of that time,
interceding for the Jews, and being sent back to attempt a reconciliation. The
letter does not call him high priest (and therefore it is just conceivable that
the identification is wrong), but it is clear that Menelaus remained in office
as high priest until put to death while in the entourage of Antiochus V Eupator in / .. Here  Macc. (:–) notes that it was proved to the king
that Menelaus was ‘‘responsible for all the evils’’ and also comments that he
had ‘‘committed many wrongs in relation to the Altar’’; but it says nothing
more specific about his previous role. Josephus is more definite:
For Lysias had advised the king to slay Menelaus, if he wished the Jews
to remain quiet and not give him any trouble; it was this man, he said,
who had been the cause of the mischief by persuading the king’s father

The Hellenistic World and Rome
to compel the Jews to abandon their fathers’ religion. Accordingly, the
king sent Menelaus to Beroea in Syria, and there had him put to death;
he had served as high priest for ten years, and had been a wicked and
impious man, who in order to have sole authority for himself had compelled his nation to violate their own laws.45
It must be emphasised that this passage of Josephus, whose account is secondary, filled with abbreviations and confusions, and separated by two and a
half centuries from the events in question, is the most concrete and specific
evidence we have for the responsibility of Menelaus (or any other ‘‘Helleniser’’) for the persecution by Antiochus. For what it is worth, like all our
other explicit evidence, it refers not to the creation of a syncretistic cult but
to a forcible suppression of Judaism.
None the less, it is true that Menelaus was given the high priesthood by
Antiochus; survived an armed challenge by his predecessor, Jason; kept his
office through the period of persecution, acting at least once as intermediary between the king and the Jews; and was succeeded after his execution by
Alcimus, or Yakim—a true descendant of Aaron, unlike himself—who was
appointed by Antiochus V Eupator, and was skilful enough to have himself
reconfirmed by Demetrius when he seized the Seleucid throne in  ..46
The fact that there was thus a continuity in the occupation of the high priesthood is reason enough to ask whether there was any continuity in the form
or object of the cult practised in the Temple before and after . For the
change to be such that we could properly call it a reform at least three conditions would have to be satisfied. That there was a continuity of personnel,
which there was, at least in the person of the high priest; that these individuals either initiated or were personally committed to the change, for which
the evidence is at best very slight; and, above all, that it continued to be
monotheistic, at the very least in the weak sense of recognising one god as
clearly supreme amongst others.
Notoriously, the further in time our sources are from these events, the
more specific they are about the cults and cult objects now established. But
this question, discussed brilliantly but speculatively by Bickermann 47 and
more dispassionately by Hengel (–), is absolutely central to the whole
issue. For unless the nature of the cult can be clearly established, discussions
. Ant. , – (Loeb trans.).
. For Alcimus/Yakim, see  Macc. :–;  Macc. :–; Josephus, Ant. , –,
–.
. Der Gott der Makkabäer, –.
The Maccabean Revolution

of what form of ideology or religious belief led to it become largely pointless.
We may leave out of account here those cult acts or other actions (like
eating pork) which were forcibly imposed on the population solely in order
to make them break the Mosaic law, and concentrate on what little is actually attested of the new cult forms in the Temple. For the reasons mentioned
above, the order of treatment must be chronological.
Firstly, Daniel , though it refers to the ‘‘abomination of desolation’’
(v. ) and to the establishment in Jerusalem of a foreign garrison worshipping a foreign god (v. ), cannot provide us with any specific information
about the cult.
Secondly,  Macc.  is much more helpful. It records the dedication (on
the orders of Antiochus) of the Temple to Zeus Olympius and of that on
Mount Gerizim to Zeus Xenius, the latter on the request of the inhabitants
(see above). Orgies involving intercourse with women (i.e., ritual prostitution?) took place in the Temple precincts. Unlawful sacrifices were made
on the Altar of Burnt Offering. The royal birthday was celebrated, as were
feasts of Dionysus. Chapter  reveals that pagan altars had been placed in
the market-place of Jerusalem and sacred precincts established.
Thirdly,  Macc.  reveals that many Israelites sacrificed to idols (v. ) and
that the king’s order involved the construction of ‘‘altars, sacred precincts,
and shrines for idols.’’ The ‘‘abomination of desolation’’ was set on the Altar,
and other altars were built in the cities of Judaea (). Verse  shows that
the ‘‘abomination’’ was a pagan altar placed on the Altar of Burnt Offering.
Fourthly, Diodorus /,  (see above), under the impression that the
Temple contained a cult statue of Moses, as Founder, records that Antiochus
sacrificed a sow to him on ‘‘the open-air altar.’’
Fifthly, Josephus, in BJ , , also relates that a sow was sacrificed on the
Altar, and more specifically, in Ant. , , that he built a pagan altar over
the Altar and sacrificed swine on it.
This evidence is all that we have from sources reasonably close in time to
the events. It is noticeable that they tend to concentrate on the new form of
sacrifice on the Altar of Burnt Offering. None makes any reference to the
Holy of Holies or to the establishment in the Temple of any pagan cult statue.
Later sources, however, do make specific assertions relating to cult statues.
Sixthly, Jerome’s Commentary on Daniel, using Porphyry’s detailed discussion of it written in the later third century .., makes a number of detailed statements about the cults imposed: Antiochus established a statue of
Zeus Olympius in the Temple (on :–); later he states that as well as the
simulacrum (image) of Zeus statues of Antiochus himself were placed in the

The Hellenistic World and Rome
Temple (on Dan. :; also on v. ), and that it was these which Daniel called
the ‘‘abomination of desolation.’’ 48 The chronicler Georgius Syncellus (
Dindorf ) also reports the establishment of a statue of Zeus Olympius in the
Temple, while Ioannes Malalas states that the Temple was dedicated to Zeus
Olympius and Athena (– Dindorf ).
Later writers thus clearly assumed that the attribution of the Temple to
Zeus will have involved a cult statue. If we use their testimony at all—which
we probably should not—we cannot avoid its implications simply by denying, as did Bickermann,49 that such statues will have been objects of cult. In
conclusion, however, it has to be admitted that we do not know whether
there was a cult statue, or more than one, in the Temple. All that we can be
certain of is that the Temple was dedicated to Zeus Olympius, the favoured
god of Antiochus, and that sacrifices of pigs were carried out on a pagan altar
constructed over the Altar of Burnt Offering.
But if the Temple was dedicated to one god, that was no more than a normal feature of paganism. It would imply some form of monotheism only if
our best evidence showed categorically that other cults were not imposed on
the Jewish population. But our most detailed evidence,  Maccabees, shows
precisely that a number of altars and sacred precincts were established, and
specifically that rites in honour of Diosysus were required. In short, what
was imposed was paganism.
Conclusion
Little of the evidence discussed above is absent from Hengel’s learned and invaluable book. What is offered as a reflection on it is a basic change of emphasis. I would argue, firstly, that the evidence shows how un-Greek in structure,
customs, observance, literary culture, language, and historical outlook the
Jewish community had remained down the earlier second century, and how
basic to it the rules reimposed by Ezra and Nehemiah had remained.50 That
. References are to the text in Corp. Christ. Lat. LXXVA (), S. Hieronymi Presbyteri
Opera . Commentarium in Danielem Libri III.
. Der Gott der Makkabäer, .
. I would thus not wish to accept the conclusions of Howard Clark Kee in his valuable and interesting review of the new Schürer in Religious Studies Review , Oct. , –:
namely that Schürer’s work, even as revised, represents an outdated conception of Jewish
life in this period as dominated by the Law; rather the future belongs with Hengel’s conception of it as profoundly Hellenised. The evidence now available seems to me to suggest
otherwise even more strongly than did that known to Schürer.
The Maccabean Revolution

there was a ‘‘Hellenising movement’’ is beyond all question, but by far the best
evidence for it relates to the first phase, the establishment of Jerusalem as a
Greek city called ‘‘Antiochia’’ and its gymnasium in the s. To see the whole
course of events as the work of a ‘‘Hellenising party’’ is to ignore the change
of the high priesthood from Jason to Menelaus; to underestimate the violence and extent of direct Seleucid administrative and military intervention;
and to build an immense structure on a few fragments of evidence (almost
entirely from Josephus’ confused account) laying the responsibility for what
happened on Menelaus and his associates.
However difficult it may be to accommodate it within our normal views
of Hellenistic rule in Asia, the evidence, both Jewish and pagan, asserts that
the programme of ‘‘forced’’ conversion to Hellenism (or paganism) was that
of Antiochus and was carried out by his agents. He did not, it is true, conduct it in person; moreover, by  .., the need to exact taxes from the
central Asian provinces of his empire was more pressing than the concerns
of Judaea, and he died far away in Elymais at just about the moment when
Jewish worship was restored in the Temple. Even before that he had begun to
compromise, and his son Antiochus V at once formally abandoned his father’s
attempt to force the Jews to accept ‘‘Greek customs.’’
That was what it had been (as indeed Hengel admits, ). It follows that
literary or philosophical identifications of the Jewish God with Zeus or a
‘‘Highest God’’ of the pagan pantheon would have only a restricted relevance,
even if—which is not the case—they could be shown to have been shared by
persons living in Judaea in the period leading up to the ‘‘reform attempt.’’ For
the events of  we lack not only any convincing evidence of the theology
of the putative ‘‘reformers’’ or (as above) for their influence on Antiochus,
but also any adequate grounds for believing that the cult then established had
clearly monotheistic features, or retained anything from the previous worship in the Temple which would entitle it to be called a ‘‘reform’’ rather than
an abolition.
These are difficult questions, if of the greatest historical importance; and
no one, least of all the author, will wish to claim that he has divined with
certainty a reality different from that portrayed in Hengel’s great work. But
it may yet be helpful to put forward, as hypotheses, a number of propositions which together would present a totally different conception of these
events from that of Hengel: (a) that what was significant about the Jewish
community of the third and second centuries .. was the superficiality of
its Hellenism; (b) that there was a reform attempt, initiated from within the
community, but confined to the high priesthood of Jason; (c) that the crisis

The Hellenistic World and Rome
of the s arose from an attempt by Antiochus Epiphanes to impose by force
the abolition of Judaism and the adoption of paganism; (d) that, while many
Jews certainly assented to this, the evidence for a Hellenising party having
instigated it is very slight; and (e) that we should not look for the intellectual
background of a syncretistic reform movement within Judaism, because we
have no clear evidence that such a movement existed.
 
Polybius between Greece and Rome *
I would like to begin with some much-quoted words of Polybius himself
(, , ). ‘‘For who is so worthless or so idle as not to wish to find out by
what steps and overcome by what sort of political structure almost all parts
of the inhabited world have, in the space of hardly fifty-three years, fallen
under the domination of the Romans, a thing which is not found ever to
have happened before?’’
The fifty-three years which Polybius refers to were those which began
in  .., the moment when, in his view, events in all parts of the Hellenised world, previously separate, began to be interconnected; they ended in
 .., when, at the battle of Pydna, Rome destroyed the first of the great
Hellenistic monarchies, that of the Antigonids, which had ruled Macedonia
for a little over a century. By ‘‘domination’’ (archē ) Polybius did not mean
what we often mean when we think of the formation of the Roman Empire:
the creation of territorial provinces and the imposition of tribute. He meant
military victory, the right to decide whether or in what form a city or a kingdom might keep its independence, and the ability to command obedience
by the threat of force.1
That same year,  .., saw the most spectacular of all examples of the
exercise of Roman domination in this sense. Antiochus IV, the ruler of the
Seleucid kingdom, based on Syria and Babylonia, had invaded Egypt, had
defeated the Ptolemies, had claimed the kingship of Egypt for himself,2 and
* First published in J. A. T. Koumoulides, ed., Greek Connections: Essays on Culture and Diplomacy (Bloomington, Ind., ), –. The translations in the text are those of the Loeb
edition.
. See P. S. Derow, ‘‘Polybius, Rome and the East,’’ JRS  (): .
. For the decisive evidence that he had, see J. D. Ray, The Archive of Hor (), .


The Hellenistic World and Rome
was outside Alexandria with his army. But at that moment there appeared
a Roman ambassador, Popillius Laenas, who handed the king the text of a
decree of the Senate, telling him to end his war with Egypt. When the king
said that he would consult with his advisers, Laenas made the famous gesture of drawing a circle in the sand round where the king was standing and
telling him to give his answer before stepping out of it. The king submitted. The earliest narrative of this famous scene comes from Polybius himself
(, ). But the humiliation of the king made an immediate impact in the
eastern Mediterranean, as we know from a pseudo-prophecy in the book of
Daniel, written only a couple of years later (:): ‘‘For the ships of Kittim
will come, and he shall be grieved and return.’’ 3
While the two Ptolemies were still in danger from Antiochus IV’s advance, the major league of Greek cities in the Peloponnese, the Achaean
league, had debated whether to send military assistance to them. Ambassadors from the two Ptolemies, brothers who were formerly at odds but now
reconciled, had arrived asking for the dispatch of , foot soldiers and 
cavalry. The infantry were to be commanded by Lycortas, Polybius’ father,
and the cavalry by Polybius himself, now probably in his thirties.4 The proposal, however, ran into difficulties; the pro-Roman party argued that all
their efforts should be directed to helping the Romans in their current war
against Perseus, the king of Macedon, in which a decisive battle was now
(rightly) expected. Polybius replied that in the previous year ( ..), when
he had been sent as ambassador to the Roman commander, he had been told
that the Romans needed no military assistance; in any case the Achaean league
could raise , or even , men if need be, so , going off to Alexandria would make no difference (, –).
In fact, the force was not sent to Alexandria. In the same year,  .., the
Romans defeated the Macedonians at Pydna, and the kingdom was dissolved.
In the following year large numbers of political figures in Greece, regarded
as anti-Roman, were taken off to exile in Rome and Italy: among these were
, from Achaea, including Polybius himself. They were to remain there
for seventeen years until their eventual release in  ..
It was in Rome that Polybius conceived his enormously ambitious plan for
a universal history which would, first, show how events in all the different
parts of the civilized (i.e., Hellenised) world came together in a set of causal
interconnections, from  .. onwards. It would therefore cover an un. On Daniel, see chapter  in the present volume.
. For the dates of Polybius’s life, see M. Dubuisson, ‘‘Sur la mort de Polybe,’’ REG 
(): .
Polybius between Greece and Rome

precedented geographical range, from Antiochus III’s campaigns in northern
India in the last decade of the third century to the Roman wars in Spain.
The work was also on an enormous scale in itself. Had it survived complete,
it would have run to over , printed pages of a Teubner text, or twenty
Loeb volumes. As it is, what remains occupies six Loeb volumes. This always
has to be remembered when we speak of what Polybius thought—or what
he seems to have omitted.
One reason the work was so long was that Polybius changed his mind
about where it should end. The original stopping point was to be the destruction of the Macedonian kingdom in  .. But at the beginning of
book  he describes why he changed his mind:
Now, if from their success or failure alone we could form an adequate
judgment of how far states or individuals are worthy of praise or blame,
I could here lay down my pen. . . . For the period of fifty-three years
finished here and the growth and advance of Roman power was now
complete. . . . But since judgments regarding either the conquerors or
the conquered based purely on performance are by no means final . . . I
must append to the history of the above period an account of the subsequent policy of the conquerors and their method of universal rule, as
well as of the various opinions and appreciations of these rulers entertained by the subjects. (, )
He therefore, in this second introduction, sketches the events which were to
occupy the last ten books (–), covering the years from  to  .. The
culminating point was to be the war of – .., in which the Achaean
league rose in revolt against Rome and was destroyed. It is crucial to his
whole historical perspective that he chose the tragic end of his own league
as his conclusion; this, along with the Roman defeat of a renewed revolt in
Macedon, was ‘‘the general disaster of all Hellas’’ (, , ).
Polybius’ second intention, in his original plan, had been not merely to
describe how all these complex events interlocked but to explain why the
Romans had been successful. The explanation, as the quotation with which
this essay begins indicated, was to be in terms of the Roman constitution or
political structure, the politeia: ‘‘Who would not wish to find out . . . what
sort of politeia had enabled the Romans to achieve domination of the whole
civilised world?’’ The reference is of course to the famous analysis of the
working of the Roman constitution and political system in book , which he
placed just after the Romans’ most crushing defeat, by Hannibal at Cannae
in  .. This in other words was the moment when, if there were weaknesses in the system, they would show up. In fact the Roman system, with its

The Hellenistic World and Rome
elaborate checks and balances, showed remarkable resilience, and Hannibal
was ultimately defeated.
As to the question of whether Polybius’ analysis of Rome was appropriate
or wholly misguided, my view is that it was not in the least inappropriate
to discuss Rome in terms of Greek political theory, or to compare it to wellknown Greek models.5 But my main point is that to understand Polybius we
have to accept that his whole perspective is that of the self-governing Greek
city, or league, of the classical and Hellenistic periods. The second point is
that Polybius’ intention, in analysing the reasons for the success of Rome,
was neutral: to give reasons for success and resilience is not in itself to recommend a system, still less to praise the results of its success.
That brings me to a third preliminary point. In going on to cover the
troubled period from  to  .., Polybius evidently did intend to introduce an element of moral judgement into his History. How in fact had the
victors used their power? There is no simple or unambiguous way of stating
Polybius’ conclusion. In a paper with the same title as this one, F. W. Walbank, the greatest modern expert on Polybius, concluded that on the whole
his view was favourable.6 I think otherwise; that Polybius, though he expresses himself obliquely, took an increasingly distant and hostile view of
Roman domination. At the very least, one point is surely remarkable by its
absence. In the whole surviving text there is not a single word to the effect
that Roman domination was a good thing or brought benefits to those who
came under it. For a man who believed that the Achaean revolution of –
 .. was a tragic error, who had spent seventeen years in Rome, and who
had friends in the highest Roman circles, this silence surely speaks volumes.
There is more to it than that, however. In the second to last of the surviving books () he comes to the Achaean war of – .., which destroyed
his own Achaean league, which he had called in book  the political system
best fitted of all for equality and freedom of speech—in fact, true democracy
(, , ). To introduce the Achaean war he puts it deliberately in a long historical perspective: going back to Xerxes’ invasion of Greece in  .., he
reviews all the major calamities and conflicts in Greek history, and concludes
that this was the greatest disaster of all; partly because it was the Achaeans’
own fault, partly because in other cases no moral blame had been incurred,
. For this view, see ‘‘The Political Character of the Classical Roman Republic (–
 ..),’’ JRS  (): – ( chapter  in F. Millar, Rome, the Greek World, and the East I:
The Roman Republic and the Augustan Revolution).
. F. W. Walbank, ‘‘Polybius between Greece and Rome,’’ Polybe (Entretiens Hardt ,
), .
Polybius between Greece and Rome

and partly because recovery from disaster had often come swiftly. This time,
by implication, no recovery was in sight (, –).
To Polybius the role of the historian was not just to record, though a history which recorded events on a universal scale gained great significance
from that alone. The historian’s role was to judge, to put events in a wider
context, and to provide lessons for the future. For that of course the historian
must himself have political and military experience; Polybius had a leading
role in a major Greek league. Second, he must be able to see the events he is
recording in perspective and, by setting them in context, to bring out their
meaning and significance. In the quotation from the first two pages of Polybius with which I began, he suggests that no one could be so indolent as not
to want to understand how Rome had achieved universal domination. The
significance of that domination would become clear, however, by comparison with the empires of the past:
The Persians for a certain period possessed a great rule and dominion,
but so often as they ventured to overstep the boundaries of Asia they
imperilled not only the security of their empire but their own existence. The Lacedaimonians, after having for many years disputed the
hegemony of Greece, at length attained it but to hold it uncontested
for some twelve years. The Macedonian rule in Europe extended but
from the Adriatic region to the Danube. . . . Subsequently, by overthrowing the Persian empire they became supreme in Asia also. But . . .
they never even made a single attempt to dispute possession of Sicily,
Sardinia, or Libya, and the most warlike nations of Western Europe
were, to speak the simple truth, unknown to them. (, )
In this last sentence he is emphasising a distinctive feature of Roman domination. It is not of course wrong to see Polybius as a historian of Rome and
its empire. Indeed when Penguin Books brought out in  a one-volume
selection of Polybius, they called it The Rise of the Roman Empire and made
the selection in such a way as to concentrate almost entirely on Roman history. However, to do that, though it has its uses, is to lose sight of Polybius’
vastly wider geographical perspective of events from India to Asia Minor,
to Syria, Egypt, Crete, and Macedonia. But above all it is to neglect the real
historical perspective which Polybius brings to the events of his age. That
is the perspective, first, of the Greek historical tradition of the classical and
Hellenistic periods. It is worth reminding ourselves that these words, ‘‘classical’’ and ‘‘Hellenistic,’’ are our terms, not his. They can, moreover, be seriously misleading. Within the past half-century books have occasionally been
published which announce that the death of Alexander the Great in  ..

The Hellenistic World and Rome
marked the end of Greek freedom, the death of the Greek city-state, and
the beginning of a new age.7 As regards the Hellenisation of the East, there
was something in this, even though that Hellenisation neither began with
Alexander nor depended entirely on his conquests.8 But that was not in any
case an aspect of history in which Polybius took much interest. As regards
mainland Greece, Hellas proper, it is nonsense.
For Polybius, as is absolutely clear, Greek history from the fifth century
to his own time was a continuum, in which there had been many disasters
but no violent break. This perspective, as we shall see, stretched back to include some mythical founders and lawgivers, like Lycurgus in Sparta, and
it of course reached to the poems of Homer. But the later archaic period,
on the evidence of the surviving text, did not play a large part in Polybius’
consciousness; nor did the monarchies of the Near East, except by way of
general allusions to the Persian Empire. His real historical starting point, or
boundary, was Xerxes’ invasion of  .. From that point Polybius’ use of
earlier history embodies an awareness of a continuous and still relevant story,
all of which was of importance for the present. For instance, when he has
narrated the Roman defeat of the Gauls in the Po Valley in the s .., he
turns immediately to Greek examples of the successful repulse of barbarians. One of them, once again, is Xerxes’ invasion; the other is the defeat of
those Gauls who invaded Greece and reached Delphi in  .., a defeat
celebrated throughout the Hellenistic world 9 (he omits to say that the victory was gained by the Aetolians, of whom he did not approve). His motives
for making these allusions are, as usual, stated with great clarity:
For indeed I consider that the writers who chronicled and handed
down to us the story of the Persian invasion of Greece and the attack of
the Gauls on Delphi have made no small contributions to the struggle
of the Hellenes for their common liberty. For there is no one whom
hosts of men or abundance of arms or vast resources could frighten
into abandoning his last hope, that is to fight to the end for his native
land, if he . . . bore in mind how many myriads of men, what determined courage and what armaments were brought to nought by the
resolve and power of those who faced the danger with intelligence and
coolness. (, , –)
. See, e.g., N. G. L. Hammond, A History of Greece to  BC, nd ed. (), .
. See, e.g., F. Millar, ‘‘The Phoenician Cities: A Case-Study of Hellenisation,’’ Proc.
Camb. Philol. Soc.  (): – ( chapter  in the present volume).
. For this, see G. Nachtergael, Les Galates en Grèce et les Sôtéria de Delphes ().
Polybius between Greece and Rome

What determines Polybius’ perspective is both a national tradition of Greek
patriotism and, as this passage illustrates, a historiographical tradition.
As it happens, the surviving text of Polybius nowhere names Herodotus, on whom of course the tradition of the defeat of Xerxes depended. The
writers of the later fifth, fourth, and third centuries, Thucydides, Xenophon,
Ephorus, Theopompus, Callisthenes, and Timaeus, to name only the major
figures, represented for him a continuous tradition of Greek history writing,
to which he constantly makes explicit reference. Indeed, his choice of the
date  .., with which he opens his preliminary narrative of Roman history, before he comes to the great conjunction of events in – .., was
determined partly by the fact that it was then that Roman forces first crossed
into Sicily; and partly by the fact that it was in that year that Timaeus’ History
had stopped (, , ; , , ). Nor was there any doubt what the proper and
central subject of a Greek history should be. This is made quite clear in what
he says of Theopompus’ History, which covered the period from  .. to
the reign of Philip of Macedon:
No one could approve of the general scheme of this writer. Having
set himself the task of writing the history of Greece from the point at
which Thucydides leaves off, just when he was approaching the battle
of Leuctra and the most brilliant period of Greek history, he abandoned
Greece and her efforts, and changing his plan decided to write the history of Philip. Surely it would have been much more dignified and
fairer to include Philip’s achievements in the history of Greece than to
include the history of Greece in that of Philip. (, , –)
As we shall see at the end, the tension between monarchic power, with its
varied threats and possible benefits, and the freedom of the Greek cities was
one of the fundamental issues which shaped Polybius’ historical consciousness. There is a very real sense in which the period of Greek history within
which Polybius most fully ‘‘belongs’’ is that which begins in the very late
fifth century, and is marked by the rise and influence of monarchs, whether
tyrants, like Dionysius I the tyrant of Syracuse (– ..), or kings, like
Philip of Macedon and his successors.10 If we allow ourselves to gain a sense
of Polybius’ historical consciousness, we might begin to feel more sympathy for the suggestion made by Hermann Bengtson that we should use the
term ‘‘Hellenistic’’ not just of the period after Alexander, but from the first
half of the fourth century onwards.11 What Polybius does not do, however,
. For this theme, see esp. J. K. Davies, Democracy and Classical Greece (), chap. .
. H. Bengtson, Griechische Geschichte 5 (), ff.

The Hellenistic World and Rome
is to treat the monarchies which played such an important part in his world
as being themselves objects worthy of the serious application of political
theory. In that he is quite explicitly the heir of the fourth-century Greek
political theory of Plato and Aristotle, to both of whom he refers repeatedly,
and whose analyses were devoted to the nuclear city-state.12 When he comes
to his famous discussion of the Roman constitution in book , the frame of
reference or comparison which he applies is first of all Sparta and the constitution given by Lycurgus (, ); later he returns to Sparta, once again,
along with the cities of Crete, Mantinea, and Carthage. Two other candidates
for consideration, Athens and Thebes, are set aside because their respective
periods of dominance, in the fifth and fourth centuries respectively, were
too brief, their political structures too unstable or fragile to be worth serious comparison with that of Rome (, –). In the end, Polybius makes
clear, what demonstrated that the Spartan constitution was also inferior to
the Roman one was the fact that it did not stand up to the strains imposed on
it by controlling a foreign empire. The Spartans had succeeded in dominating their neighbours in the Peloponnese, the Messenians. But when it came,
in the early fourth century, to a general domination of Greece, they found
themselves compelled to be dependent on the very Persians whom they had
just defeated. For they needed to ask them for the funds which their own
laws, laid down by Lycurgus, prevented them from generating at home. So
they made the inglorious Peace of Antalcidas (the ‘‘King’s Peace’’) of  ..,
in which they betrayed the freedom of the Greeks to Persia in return for financial support. By comparison, the Romans, having sought at first only the
domination of Italy, could command the resources to control an overseas
empire as well (, –).
In examining the ability of a city-state to support an empire, Polybius extends the range of political thought beyond that inherited from Plato and
Aristotle. So he does also in his marvellous account in book  (–) of the
history of his own Achaean league from its legendary origins, to its overshadowing by Sparta and then Macedon in the fourth century, and its reemergence, recovery of freedom, and rapid extension in the third. Plato and
Aristotle had not given any attention to analysing the nature of federal states.
Polybius did. By his own time, he says, the Achaean league had almost become a single city-state: ‘‘They have the same laws, weights, measures and
coinage, as well as the same magistrates, councillors and juries, and the whole
. For his explicit allusions to Plato and Aristotle, see the Teubner ed. by Th. BüttnerWobst, ,  (Aristotle) and  (Plato).
Polybius between Greece and Rome

Peloponnese only falls short of being a single city in the fact of its inhabitants
not being enclosed by one wall’’ (, , –).
As has often been noted, Polybius discusses Rome also in terms of a nuclear
city-state and did not devote any attention, so far as we know from the surviving text, to what moderns sometimes call the ‘‘Italian confederation.’’ But
in that he was in fact right, for the structure of Roman Italy was that of a set
of alliances between Rome and individual city-states and peoples, and was no
sort of league of confederation like the Achaean league.13 Where he might
have extended the range of political analysis beyond that inherited from Plato
and Aristotle was in relation to monarchy. For while the resilience of Rome
in the face of three major defeats by Hannibal in the years – .. rightly
impressed him, he might well have been struck, by comparison, by the fragility of the great Hellenistic kingdoms, which would admit defeat in a war
after the loss of a single battle, as Macedon did in  and  .., or two
battles, as with those of  and  .. by which the Romans drove Antiochus III first out of Greece and then out of Asia Minor.
Perhaps we should not blame Polybius for this oversight, however, for
even the most brilliant of modern accounts, say by Elias Bickermann or
Claire Préaux,14 have hardly succeeded in explaining what a Hellenistic monarchy was really like as a system, or how it held together at all. Confronted
with monarchies, Polybius turns either to explanations in terms of personal
character or to observations on the instability of human fortune. In both
approaches, once again, the framework is supplied by the earlier history of
Greece. For instance, when he wants to set in context the fact that Philip V
of Macedon in  .. allowed his army to destroy colonnades and statues
in the Aetolian city of Thermon, he reflects on the models which the king
should have followed: Philip II’s clemency to the Athenians after the battle of
Chaeronea in  ..; Alexander’s care not to destroy the temples of Thebes
in Boeotia in  .., or his preservation of temples in the Persian Empire
(, ). But Philip V, he says, was either too young or had the wrong advisers,
and lost the chance of gaining the good reputation which mercy would have
won him (, –). Later Polybius represents Philip V, now much older and
perhaps wiser, telling his sons to read tragedies, myths, and histories, and to
think of the disastrous effects of strife between brothers; or, alternatively,
to think of the kings of Sparta, whose success had been gained by mutual
. See F. Millar (n. ).
. E. J. Bickermann, Institutions des Séleucides (); C. Préaux, Le monde hellénistique I–II
(), –.

The Hellenistic World and Rome
concord and obedience to the laws and the ephors (the five annually elected
magistrates). There was also the contemporary example of the two brothers,
Eumenes II and Attalus II of Pergamon, whose concord had raised their kingdoms to greatness from small beginnings (, ).
When he comes to the great event with which I began, the Roman defeat of Perseus, king of Macedon, at Pydna in  .., and to the end of the
Macedonian monarchy, Polybius turns back again to the fourth century and
quotes a marvellous passage from Demetrius of Phalerum, the philosopher
who was the effective ruler of Athens for a decade in the very early Hellenistic period, from  to  .. Demetrius in his work On Fortune had
reflected on the sudden end of the Persian Empire (, ).
Do you think, that fifty years ago either the Persians and the Persian
King or the Macedonians and the King of Macedon, if some god had
foretold the future to them, would ever have believed that at the time
when we live the very name of the Persians would have perished utterly
—the Persians who were masters of almost the whole world—and that
the Macedonians, whose name was formerly almost unknown, would
now be lords of it all? But nevertheless Fortune . . . now also . . . makes it
clear to all men, by endowing the Macedonians with the whole wealth
of Persia, that she has but lent them these blessings until she decides to
deal differently with them.
A century and a half later, as Polybius goes on to reflect, this inspired
prophecy had come true, and Fortune had withdrawn from Macedon the
blessings which she had briefly given.
Such reflections were of course not unique to Polybius. By his time, they
had entered Roman culture as well. So, as he himself records (, , ), when
the Roman general Scipio Aemilianus witnessed the destruction of Carthage
in  .., he recalled the fate of the empires of Assyria, Media, Persia, and
Macedon itself, and quoted two lines from the sixth book of Homer’s Iliad
(–):
A day will come when sacred Troy shall perish,
And Priam and his people shall be slain.
For Polybius the uses of history did not always need to be on so high a philosophical level as that. Writing for a Greek audience, to record and explain
the rise of Rome to universal domination, he could on occasion use crossreferences to Greek history purely for dating purposes, that is, to anchor
events in earlier history in a context which was, or was supposed to be, famil-
Polybius between Greece and Rome

iar to his educated readers. It is worth stressing how detailed a familiarity
with Greek history he seems to presume. ‘‘Who had not read,’’ he says at one
point, ‘‘that when the Athenians, in conjunction with the Thebans, entered
on a war against the Lacedaimonians, sending out a force of ten thousand
men and manning a hundred triremes, they decided to meet the expenses by
a property-tax, and made a valuation for this purpose of the whole of Attica
including the houses and other property?’’ (, , –). Whether his audience really did have an immediate recall of this episode or not, it belonged,
once again, in the fourth century, to be precise in  .., at the moment
of the foundation of the Second Athenian Confederacy.
When he applies Greek history to Roman history for dating purposes,
he normally selects points of reference which were more familiar. So, for
instance, the year of the earliest treaty between Rome and Carthage is identified not only as the first year of the annual pair of consuls at Rome, and the
year after the expulsion of the Roman kings; it was also ‘‘twenty eight years
before the crossing of Xerxes into Greece’’ (, , –)—that is, in our terms,
 .. Or again, when the Gauls captured all of Rome except the Capitol,
it was the nineteenth year after the battle of Aegospotamoi (which ended
the Peloponnesian War, in  ..), the sixteenth year before the battle of
Leuktra (in  ..), the same year as the Peace of Antalcidas with Persia,
and also the same year as that in which Dionysius I, the tyrant of Syracuse,
was besieging Rhegium,  .. (, , ). These multiple references can only
have been helpful to an audience to whom the details of fourth-century history were familiar. Thanks to the Sicilian historian Timaeus, Polybius’ Greek
history in the fourth and third centuries also embraced the history of Sicily.
One of the many revelations of his history, if we think of it as a Greek history, is that it gives us something rather rare, a perspective which is decidedly
non-Athenian, and is instead, firstly, Peloponnesian and, secondly, Sicilian.
In the third century too, he can use the same anchoring device, relating
the Roman victory at Lake Vadimon in  .. both to Pyrrhus’ crossing
into southern Italy in  .. and to the defeat of the Gauls at Delphi in
 .. (, , ).
More significant, however, was the use of examples from Greek history
as points of reference for historical and political judgements. Once again,
such examples can only be useful if they are familiar and are charged with
some meaning for their readers. For example, it might be right, Polybius
suggests, to see Hannibal in the light of persons, or whole peoples, whom
circumstances had caused to act variably, or contrary to their real character. For instance, Agathocles, tyrant of Syracuse in – .., had been

The Hellenistic World and Rome
by turns cruel and benevolent; Cleomenes, king of Sparta in the third century (– ..), exhibited similar contradictions; the Athenian people
behaved differently under the leadership of Aristides and Pericles than they
did under Cleon and Chares; so also did the Spartans, under kings Cleombrotus and Agesilaus, in the early fourth century (, ). Again, the examples
extend from the fifth century into the Hellenistic period. Examples derived
from Herodotus play no part in Polybius’ mental framework, at least in the
surviving text, which is less than one-third of the original.
What Polybius does use from archaic or legendary Greek history does not
derive from Herodotus but rather from separate traditions about early Sparta
and its lawgiver, Lycurgus. It is these which he applies to Rome and its constitution in book , and these which he also uses in interpreting the role of
the great Roman general, Scipio Africanus, commander in Spain against the
Carthaginians from  .. onwards:
To me it seems that the character and principles of Scipio much resembled those of Lycurgus, the Lacedaimonian legislator. For neither
must we suppose that Lycurgus drew up the constitution of Sparta
under the influence of superstition and solely prompted by the Pythia,
nor that Scipio won such an empire for his country by following the
suggestions of dreams and omens. But . . . Lycurgus made his own
scheme more acceptable and more easily believed by invoking the
oracles of the Pythia in support of projects due to himself, while Scipio
similarly made the men under his command more sanguine and more
ready to face perilous enterprises by instilling into them the belief that
his projects were divinely inspired. (, , –)
But the center of Polybius’ historical culture was the experience of the Greek
city-states from the fifth century onwards, and above all from the fourth
century onwards, the period of the rise and intrusion of monarchic power.
Thus, when dealing with the complex and catastrophic events of  and
 .., which culminated in the more or less simultaneous destruction by
the Romans of Carthage and Corinth, he defends his procedure in taking his
narrative backwards and forwards between the two theatres of war by reference to existing histories of fourth-century Greece: ‘‘When dealing with
Thessalian affairs and the exploits of Alexander of Pherae, they [these historians] interrupt the narrative to tell us of the projects of the Lacedaimonians
in the Peloponnese or of those of the Athenians and of what happened in
Macedonia or Illyria, and after entertaining us so tell us of the expedition of
Iphicrates to Egypt and the excesses committed by Clearchus of Pontus’’ (,
, –). Once again, these events, mainly of the s and s .., are assumed
Polybius between Greece and Rome

to be familiar to his readers. The same historical background had also been
deployed when Polybius first came to the issue of how the Romans acted in
the final war against Carthage, and how their actions should be judged.
Polybius conceals his own opinion by the device of setting out four different opinions of Rome’s conduct which were held in Greece. It should be
emphasised, however, that none of the four opinions quoted is positively
favourable: of the two more favourable, one held that it was a sign of prudence on the part of the Romans to destroy an ancient enemy when the opportunity offered; and the other maintained that the Romans had committed
no actual offense in international law. Of the two unfavourable views, one
held that the level of deceit used by the Romans amounted to impiety and a
breach of treaty obligations. The other was that the Romans had previously
taken their warfare only to the point of forcing their opponents to submit to
their orders. But now they were deserting their former principles in favour
of a lust for domination ( philarchia) like that of Athens and Sparta and would
come to the same bad end (, ). Here, too, the history of the fifth and fourth
centuries is recalled to the reader.
But the most remarkable and detailed of all the occasions where Polybius
makes use of the earlier history of Greece belongs at a previous stage in his
narrative; very significantly this is at exactly the point where Roman military force first became a major factor in the life of mainland Greece. In 
or  .. the Romans, threatened by an alliance between Philip V of Macedon and Hannibal, themselves made an alliance with the Aetolian league,
the major power in north-western Greece.15 Then in  .. ambassadors
from Acarnania, which was in alliance with Philip, and from Aetolia, now
allied to Rome, presented themselves simultaneously at Sparta, each hoping
to persuade the Spartans to join his side. To underline the significance of
the occasion, Polybius in book  gives each of the two ambassadors a speech
which presents a view of the historic role of the Macedonian monarchy in
Greece and the light in which the Roman intervention should be viewed.
(There is no way of saying how far the speeches in Polybius resemble anything that was actually said on this occasion, shortly before his birth. What
matters is simply what we have in the text, two speeches designed to bring
out the significance of a major turning point.)16
The Aetolian speaks first and, as so often in the text of Polybius, goes
back to the mid-fourth century, to the capture of Olynthus by Philip II and
his suppression of Sparta; then Alexander’s destruction of Thebes, and Anti. H. H. Schmitt, Die Staatsverträge des Altertums III (), no. .
. See esp. F. W. Walbank, ‘‘Polybius and Rome’s Eastern Policy,’’ JRS  (): .

The Hellenistic World and Rome
pater’s victory in the Lamian War of  .. He then comes to the various Successors who ruled in Greece from the late fourth century onwards:
‘‘And who is ignorant of the actions of Cassander, Demetrius and Antigonus
Gonatas, all so recent that the memory of them is quite vivid? Some of them
by introducing garrisons to cities and others by introducing tyrannies left no
city with the right to call itself unenslaved.’’ Finally he comes to the acts of
violence which Philip V himself had committed in Greece. In the speech as
preserved, apparently almost complete, there is no reference to the Romans,
except to say that with their aid Philip was likely to be defeated (, –).
Whatever the original speaker in  .. really said, Polybius could have
used this point in his narrative to say something positive about the potential
Roman role in Greece. He does not. The real issue, as always, lay elsewhere:
the preservation of the freedom of the Greek cities in the face of the threats
posed by successive kings and dynasties.
Then the Acarnanian ambassador gives a speech which reviews the same
historical period, but from the opposite point of view. Philip II, he says, by
defeating the Phocians in the Sacred War, that is, in  .., had saved the
liberty of Greece. In the Peloponnese Philip had come in response to appeals
and had used his power to arbitrate between Sparta and its enemies. His son
Alexander had punished the Persians for their offenses against Greece and ‘‘in
a word he made Asia subject to Greece.’’ As for later crimes in Greece, it was
the Aetolians who were most guilty. It was the Macedonian monarchy of the
Antigonids, which ruled from  .. onwards, which had protected Greece
against the northern barbarians. Even Antigonus Doson, king of Macedon,
in defeating Sparta it in  .., had done so in order to liberate it from a
tyrant, namely King Cleomenes.
But now, the Acarnanian ambassador goes on, it is no longer a matter of
alliances between Greeks: ‘‘But now Greece is threatened with a war against
men of a foreign race [the Romans] who intend to enslave her, men whom
you [the Aetolians] fancy you are calling in against Philip, but whom you
are really calling in against yourselves and the whole of Greece.’’ All Greeks
should beware, especially the Spartans who had once thrown into a well the
ambassador sent by Xerxes to demand submission and had sent Leonidas to
defend the liberty of Greece at Thermopylae. The Romans had already committed atrocities in Greece: ‘‘A fine alliance this for anyone to determine to
join, and especially for you Lacedaimonians, who, when you conquered the
barbarians [i.e., the Persians at Plataea in  ..], decreed that the Thebans
were to pay a tithe to the gods for having decided under compulsion, but
alone among the Greeks, to remain neutral during the Persian invasion’’ (,
–).
Polybius between Greece and Rome

Once again, whatever the real Acarnanian speaker of  .. actually said,
Polybius certainly had the freedom to select what to put in his own narrative,
perhaps even the freedom to invent appropriate words. It is surely significant
that at the moment of the first substantial Roman involvement in Greece, he
makes a speaker represent them as foreigners intent on enslaving Greece, directly comparable to the Persians, those barbaroi whose defeat was the central
event in Greek history.
What I want to stress, however, is not the implicit reservations in Polybius’ attitude to Rome; though nothing could be more false, in my view,
than the idea that, in explaining to the Greek world how and why Rome had
gained universal domination, he was also recommending, or even defending, Roman rule. What is important is the fact that Polybius’ History really is
the product of his earlier experience as a central figure in the self-governing
Achaean league of cities which occupied a large part of the Peloponnese. To
Polybius that part of the past which mattered, that past from which lessons
could be drawn, was the experience of the Greek city-states since the victory
over Xerxes. It was a continuous history, all of which offered lessons and examples that were relevant to the present. Polybius would have been surprised
to learn that something called the Hellenistic age had begun in  .., and
that he himself was a Hellenistic historian. He would surely have supposed
that he was simply a Greek one.
 
The Greek City in the Roman Period *
This paper will concentrate on the imperial period, the first three centuries .., when ‘‘the Greek city’’ is more visible to us than at any other time.
For it is from this period that the vast majority of the surviving remains of
Greek cities date; it was in these centuries, except for the last few decades,
that the largest number of Greek cities struck coins; and, above all, it was in
this period that the Greek cities provided the fullest expression of their own
communal identity, through the medium of inscriptions. Although there
are cities, such as Athens, Ephesus, Miletus, and others in Caria and Lycia,
which provide substantial numbers of Hellenistic inscriptions, in almost all
cases there is a sharp decrease in the extremely troubled period of the first
century .., the time of the Mithridatic wars and the Roman civil wars,
largely fought on Greek soil. The victory of ‘‘Imperator Caesar Divi filius’’
(that is, ‘‘Imperator Caesar son of the deified [ Julius]’’), soon to add the name
‘‘Augustus,’’ both allowed and stimulated the production of public inscriptions on an unprecedented scale. In this sense the Augustan regime marks
an epoch in the history of Greek inscriptions which is almost comparable to
that which it certainly signals, as Géza Alföldy has shown, in the history of
Latin epigraphy.1 In the inscriptions, as in the coins and in the physical and
monumental structure of the Greek cities, the person, that is, the name and
the image of the emperor, was to play an essential role.2 How Greek cities ex* First published in M. H. Hansen, ed., The Ancient Greek City-State (Copenhagen, ),
–. This paper owes a great deal to the comments of the participants in the colloquium,
above all those of Ph. Gautier.
. G. Alföldy, ‘‘Augustus und die Inschriften: Tradition und Innovation,’’ Gymnasium 
(): .
. For a brief sketch of some aspects of this transformation, see F. Millar, ‘‘State and

The Greek City in the Roman Period

pressed their complex relationships to the new line of individual rulers is one
of the central aspects of what ‘‘the Greek city’’ of the imperial period was.
In a wider sense, however, the expression ‘‘the Roman period,’’ as regards
the history of Greek cities, covers a far larger time scale, which begins centuries earlier and continues several centuries later than the High Empire.
The Greek expansion into Italy, Sicily, and the coasts of Gaul and Spain in
the archaic period had meant that from the very beginning Rome belonged
on the fringes of the Greek world. Caere (Agylla), a few kilometres from
Rome, was adopting the custom of the Greek agōn (contest), as a form of expiation, in the middle of the sixth century.3 At the end of that century, as
a detailed Greek narrative preserved by Dionysius of Halicarnassus records,
the tyrant of Cumae, Aristodemus, played an important role at the time of
the expulsion of Roman kings and in conflicts with the Etruscans.4 Similarly,
Roman contacts with Massilia go back at least to the early fourth century,
when spoils from the Gauls were deposited by the Romans in the Massaliote
treasury at Delphi, and when Justin claims that there was already a treaty between Massilia and Rome.5 Already in the next couple of centuries, Greek
cities were presented with the problem of how to construe the identity of
the Romans, the significance of their claimed descent from Aeneas, and thus
the nature of their relationship to themselves. These questions are, of course,
already explicit in the famous inscription of the s .. from Lampsacus,
describing an embassy to Massilia and then to Rome, to claim protection
(against Antiochus III) on the basis of a mythical common descent.6 Other
elements of the mythical identity of Rome are reflected in a more recently
found inscription of about the same date from Chios, recording the creation
there of a monument representing ‘‘the founder of Rome, Romulus, and his
brother Remus.’’ 7
By that time of course all the Greek cites of Italy had become socii navales
(allies providing Rome with warships) of Rome, while one, Posidonia, had
Subject: The Impact of Monarchy,’’ in F. Millar and E. Segal, eds., Caesar Augustus: Seven
Aspects (), – ( chapter  of F. Millar, Rome, the Greek World, and the East I: The
Roman Republic and the Augustan Revolution), which owes much to S. R. F. Price, Rituals and
Power: The Roman Imperial Cult in Asia Minor ().
. Herodotus , .
. Dionysius, Ant. Rom. , –.
. Plutarch, Camillus ; Justin , , .
. Syll.3 no. . English translation in M. M. Austin, The Hellenistic World from Alexander
to the Roman Conquest (), no. .
. See P. S. Derow and W. G. Forrest, ‘‘An Inscription from Chios,’’ ABSA  (): 
SEG XXXI, no. . See O. Hansen, Eranos  (): .

The Hellenistic World and Rome
been re-founded as the Latin colony (colonia) of Paestum, and another, Dicaearchia, as the Roman colony of Puteoli.8 The Greek cities of Sicily had
provided the battleground for the First Punic War;9 while in the Second the
kingdom of Hiero had come to an end, Syracuse had fallen, and the whole
island had become a Roman province, with a complex network of statuses
and obligations, later to be illuminated by Cicero’s speeches against Verres.10
The history of the later Greek city under Roman rule in the West—on the
Mediterranean coast of Spain, on the coast of Gaul (Massilia with Nicaea
and Antipolis), in Italy (above all Neapolis, still a major centre for Greek
festivals in the imperial period), and in Sicily—is a major historical topic,
which cannot be properly treated here. It need only be stressed, as regards
the complex relations of the wider Greek world to Rome in the Hellenistic
period, that this area, though certainly marginal, was never unknown or irrelevant. In Livy’s narrative of the year  .., for instance, we read how
a meeting of the Aetolian league was addressed by ambassadors from Macedon, who reminded the league that Syracuse, Messina, Lilybaeum, Regium,
Tarentum, and Capua were all now subject to Rome.11 By the time that Livy
was writing Diodorus had already given the history of Greek Sicily, from
the earliest times to his own day, a central place in his universal Bibliotheca;12
while if any Greeks had ever been disposed to read Latin, they could have
studied the most ‘‘Hellenistic’’ of all universal histories, the Historiae Philippicae of Pompeius Trogus, a Vocontian who surely owed his historical culture
to the influence of Massilia.13
This is not the place to rehearse the conclusive steps which gave Rome a
central place in the Greek world: first, the suppression of the revolt of Andriskos, the defeat of the Achaean league, and the establishment of the province
of Macedonia, including Greece proper; and, then, the end of the Attalid
kingdom and the formation of the province of Asia. But for how a Greek city
would manage its dealings both with Roman governors and with the politi. For a general account of Posidonia/Paestum, as revealed by excavation, see J. G.
Pedley, Paestum: Greeks and Romans in Southern Italy (); for Dicaearchia/Puteoli, M. Frederiksen, Campania (), chap. .
. D. Roussel, Les Siciliens entre les Romains et les Carthaginois à l’époque de la première guerre
punique ().
. For a survey of later Republican Sicily, see R. J. A. Wilson, Sicily under the Roman
Empire: The Archaeology of a Roman Province,  ..–..  (), chap. : ‘‘Background.’’
. Livy , –.
. See K. S. Sacks, Diodorus Siculus and the First Century (), esp. –.
. See J. M. Alonso-Nuñez, ‘‘An Augustan World History: The Historiae Philippicae of
Pompeius Trogus,’’ Greece and Rome  (): .
The Greek City in the Roman Period

cal institutions of the city, a central place will now be claimed by the two vast
inscriptions from Claros, published by Louis Robert (posthumously) and by
Jeanne Robert.14 These two inscriptions, neither complete, of the last third of
the second century, still come to  lines (that for Polemaios) and  lines
(for Menippos) and provide by far the most complex and detailed picture so
far available of Greek cities in the period of the transfer of power from the
Attalids of Pergamum to Rome. Out of many details, only one need be noted
here: that by gaining the friendship of leading Romans Polemaios was able
to benefit his fellow citizens (of Colophon) by creating for his native city
relations of patronage ( patrōneiai, in the plural) with the ‘‘best men.’’
As regards the first century .., we should not forget the small Greek
cities which lined the coast of Illyricum, for instance, Issa and its colony Tragurion, whose embassy to Julius Caesar at Aquileia is recorded in an inscription.15 Or, on the opposite shore of the Mediterranean, the Greek cities of
Cyrenaica, which were abandoned to fend for themselves for several decades
after Ptolemy Apion had left his kingdom to the Roman people in  ..
This interval of involuntary independence is now brilliantly illuminated by
an inscription from Berenice: after the death of the king the city was first
besieged by kakourgoi (wrongdoers), and then, being unwalled, was twice
sacked by pirate fleets. Very typically, the inscription is in fact in honour of
a local benefactor, Apollodorus, who apart from his military role seems also
to have gone on an embassy to the Roman Senate.16
A very similar situation, of acute political and military danger, necessitating political and military action to seek protection, or at least benevolence, from wherever it could be found, is reflected in the famous inscription
from Dionysopolis on the Black Sea honouring a citizen named Acornion; he
had performed ceremonials during the stay there over the winter (probably
/ ..) of C. Antonius, had been on an embassy to Burebista, ‘‘first and
greatest of the kings in Thrace,’’ in the interests of his city, and had also been
sent by Burebista as ambassador to Pompeius at Heraklea Lynkestis. There
. L. Robert and J. Robert, Claros I: décrets hellénistiques (). The passage quoted is
from the decree for Polemaios, col. II, lines –. See also J. Touloumakos, ‘‘Zum römischen Gemeindepatronat im griechischen Osten,’’ Hermes  (): , and esp. the important paper by J.-L. Ferrary, ‘‘Le statut des cités libres dans l’Empire Romain à la lumière
des inscriptions de Claros,’’ CRAI (): .
. See R. K. Sherk, Roman Documents from the Greek East: Senatus Consulta and Epistulae
to the Age of Augustus (), no. .
. Edited by J. M. Reynolds in J. A. Lloyd, Excavations at Sidi Khrebish, Benghazi (Berenice)
I (supp. to Libya Antiqua , ), , no. . See A. Laronde, Cyrène et la Libye hellénistique
(), – (photograph, text, translation, and discussion).

The Hellenistic World and Rome
he had taken the opportunity to represent the interests not only of the king,
but also of his city. As its benefactor, in a way which was becoming a prime
function of the communal institutions of the Greek city, he was to be honoured with a crown each year at the Dionysia and with a statue at the most
conspicuous point in the market-place (agora).17
The story could be paralleled many times over in the complex, disturbed,
and violent relations of the cities of the Black Sea region with local kings
and dynasts, with Mithridates, and with the Romans. But if we are to think
of what the ‘‘Roman peace’’ brought by Augustus really meant, one essential
starting point is the whole succession of ‘‘local histories’’ offered by the Geography of Strabo. For obvious reasons of local knowledge and sympathy these
gain an added force when they relate to northern Asia Minor: not only his
wonderful evocation of his own city, Amaseia in Pontus (), but also the
brief vignette of the rise of Gordioucome to be the city of ‘‘Iuliopolis’’ (),
or his account of Sinope in the later Hellenistic and Republican period:
The city itself is beautifully walled, and is also splendidly adorned with
gymnasium and market-place and colonnades. But although it was such
a city, still it was twice captured, first by Pharnaces, who unexpectedly attacked it all of a sudden, and later by Lucullus and by the tyrant
who was garrisoned within it, being besieged both inside and outside
at the same time; for, since Bacchides, who had been set up by the king
as commander of the garrison, was always suspecting treason from the
people inside, and was causing many outrages and murders, he made
the people, who were unable either nobly to defend themselves or to
submit by compromise, lose all heart for either course. At any rate,
the city was captured; and though Lucullus kept intact the rest of the
city’s adornments, he took away the globe of Billarus and the work
of Sthenis, the statue of Autolycus, whom they regarded as founder of
their city and honoured as god. The city had also an oracle of Autolycus. He is thought to have been one of those who went on the voyage
with Jason and to have taken possession of this place. Then later the
Milesians, seeing the natural advantages of the place and the weakness
of its inhabitants, appropriated it to themselves and sent forth colonists
to it. But at present it has received also a colony of Romans; and a part
of the city and the territory belong to these.18
. Syll.3, no.  IGR I, no.  G. Mihailov, IG Bulg. I 2 (), no. . English
translation in R. K. Sherk, Rome and the Greek East to the Death of Augustus (), no. .
. Strabo, Geog. , ,  (), Loeb trans. (with minor emendations).
The Greek City in the Roman Period

Nothing could more accurately convey the sense of a long tradition, a period
of acute danger and disturbance, the value placed on public buildings, or (in
this case) the impact of Roman colonisation.
Equally significant, if sometimes more confused geographically, is Strabo’s
account of the Phoenician cities: their extremely varied fortunes during the
progressive break-up of the Seleucid kingdom; the rise here too, as in Syria
proper, of a generation of local tyrants, finally suppressed by the Romans;
and the imposition of the Roman peace.19 In the case of Phoenicia Strabo
makes a direct and unmistakable connection between the disorders of the
first century .. and the establishment in  .. of the colony of Berytus—
Colonia Iulia Augusta Felix Berytus, which was to be perhaps the only established ‘‘island’’ of Latin culture in the whole of the Greek world of the eastern
Mediterranean.20
In spite of all the problems which its composition raises, the Geography of
Strabo cannot but serve as the most important picture which we have of the
Greek world as it was when imperial rule began. Precisely two of its most
important functions are, firstly, to recall how the present situation of each
place could be seen in the context of a distant, often mythological, past; and,
secondly, to record how so many places had faced violent external threats
and internal disturbances in the first two-thirds of the first century .. In
the established Empire it is in general true that cities, when they collectively
recalled the past, tended to avert their gaze from the Hellenistic and early Roman periods, to focus on mythological origins and (if they could) on the classical past.21 Strabo is closer in time to the most troubled period, and therefore
all the better evidence for how much the imposition of peace really meant.
One place of which Strabo makes only two passing mentions is Aphrodisias in Caria.22 It ought to have attracted his attention, however, for it too
had taken part in the military conflicts of the first century. In  .., when
the two communities of the ‘‘Plaraseans’’ and the ‘‘Aphrodisieis’’ still formed
. For the political history of the Greek cities of both Phoenicia and Syria proper, see
the very useful studies by J. G. Grainger, The Cities of Seleukid Syria (), and Hellenistic
Phoenicia ().
. Strabo, Geog. , ,  (). On Berytus as a colonia, see F. Millar, ‘‘The Roman Coloniae of the Near East: A Study of Cultural Relations,’’ in H. Solin and M. Kajava, eds., Roman
Eastern Policy and Other Studies in Roman History (), – ( chapter  of the present
volume).
. E. Bowie, ‘‘Greeks and Their Past in the Second Sophistic,’’ Past and Present  ():
 M. I. Finley, ed., Studies in Ancient Society (), .
. Strabo, Geog. , ,  (); , ,  ().

The Hellenistic World and Rome
a joint political unit, their people (dēmos) had decided to march out, accompanied by paroikoi and slaves, to help Q. Oppius, besieged by Mithridates in
Laodicea. Whether or not they had ever arrived (or had contrived not to arrive in time?), they had taken care, when Oppius was released after the Peace
of Dardanus, to remind him of their loyalty, and to seek his patronage ( patrōnēa), which he granted; he would also, he wrote, speak in their favour before
the Senate and People when he reached Rome. In the Triumviral period the
city gained both a treaty (orkion) and a long decree of the Senate (senatus consultum), of  .., with provisions for its free status. Everything was owed,
as is quite clear, to the identification of the Aphrodite whose temple stood
there with Venus, as the mythological ancestress of the Julian house. But the
locus of power in Rome was changing. When the new ruler wrote (whether
before or after  .. is uncertain) to tell the Samians that the privilege
given to the Aphrodisians was unique, he said ‘‘You yourselves can see that
I have given the privilege of freedom to no people except the Aphrodisians.’’ 23 Such rights could now be seen as being at the personal disposition of
a monarch.
These inscriptions are of crucial importance for understanding the Greek
city of the imperial period. Firstly, as inscriptions, they are in fact of the
imperial period. The two relating to Oppius were cut in the second century ..; the long series from the Triumviral period belong to the ‘‘archive
wall’’ on the parodos (entry passage) of the theatre and date, as inscriptions,
to the first half of the third century. Aphrodisias had in fact no classical, still
less archaic or mythological, history, but had emerged as a privileged community in the last decades of the Republic. Here at least, that troubled period
was not forgotten, and could not be. How significant was it, for a Greek city
of the second century .., that such a re-cut inscription might serve to remind its citizens of one essential role of every city community of the archaic,
classical, and Hellenistic period which it had now lost, namely its military
function? In that respect, even if Platea might send off a contingent to assist
Marcus Aurelius on his northern campaigns,24 there could be no continuity
with the Greek city of before the Augustan peace. Or might there be, none
the less? The poorly recorded history of the third-century invasions shows
that some cities might indeed resume their ancient military role.25
. J. M. Reynolds, Aphrodisias and Rome (), nos. – (Oppius); – (Triumviral
period). The phrase quoted is from no. .
. A. Plassart, ‘‘Une levée de volontaires thespiens sous Marc-Aurèle,’’ Mélanges
G. Glotz II (), ; the correct context was established by C. P. Jones, ‘‘The Levy at
Thespiae under Marcus Aurelius,’’ GRBS  (): .
. See F. Millar, ‘‘P. Herennius Dexippus: The Greek World and the Third-Century
The Greek City in the Roman Period

Equally significant, the archive from Aphrodisias is one reflection of the
range of privileged statuses which emerged out of the conflicts of the late
republican period, when cities sought the favour of Rome, while Rome,
equally, used all possible diplomatic means to alleviate the effects of brutal
oppression and exaction by grants of alliances and favours. The cities of the
Greek world in the imperial period might therefore retain one or more of
the following rights. (It would be extremely difficult to state what combinations of these rights were normal, or how common it was for one to be
enjoyed without the others.) The possibilities were: a treaty, as enjoyed by
Aphrodisias and also (for instance) by Tyre, as Ulpian’s well-known celebration of it recalls;26 libertas (freedom), meaning, as the imperial documents
from Aphrodisias put it, that the city was ‘‘exempt from the typos [formula]
of the province.’’ This certainly implied exemption from the jurisdiction,
and the personal visits, of the governor; but it does also seem, as is shown
in Hadrian’s letter to Aphrodisias, to have carried with it exemption (immunitas) from Roman taxation. It is not, however, possible to assert that these
two latter rights could never be dissociated.27
Finally, there was the status of Roman colony, which almost certainly, in
this early period, carried with it automatically exemption from all forms of
direct taxation, what the lawyers were later to label tributum soli (land tax)
and tributum capitis (poll tax). At the moment of Actium, Roman colonisation was still a minor phenomenon in the Greek world. The known cases are
Tauromenium, probably in  ..;28 Corinth, re-founded in  ..; Philippi,
made a colony in  ..; Cassandrea; Buthrotum; probably Dium; Lampsacus; possibly Alexandria Troas and Parium; Apamea in Bithynia; briefly
Heraclea Pontica; and Sinope (see above).
There are many uncertainties in the list, which would be considerably
lengthened if we added the certainly Augustan colonies. The most notable of
these are, firstly, a group in Sicily; secondly, Berytus (see above); and, finally,
the important series in Pisidia and neighbouring regions.29 We will look later
Invasions,’’ JRS  ():  ( chapter  of F. Millar, Rome, the Greek World, and the East II:
Government, Society, and Culture in the Roman Empire).
. Dig. , , .
. Reynolds, Aphrodisias and Rome, no.  (Hadrian);  (legal prevention of visits by
proconsul). For discussions of these complex questions, see R. Bernhardt, Imperium und Elutheria: Die römische Politik gegenüber den freien Städten des griechischen Ostens (); Polis und
römische Herrschaft in der späten Republik, – vor Chr. (); and Ferrary (n. ).
. So Diodorus , , . See Wilson (n. ), – (suggesting that the actual colonisation
took place in  ..).
. See still B. M. Levick, Roman Colonies in Southern Asia Minor (): Antiochia in

The Hellenistic World and Rome
at the very significant process, characteristic of the established Empire, by
which emperors came to grant the title of colony to Greek cities, without
the settlement of actual colonists; for even this involved a transformation of
the formal structure and constitution of each city, and (in theory at least) the
public use of Latin.30
The ‘‘real’’ colonisation of the Caesarian, Triumviral, and Augustan period
(the only major period of colonisation outside Italy in Roman history) should
be stressed, however, for it combined with two other very well-known processes to produce marked changes in what we understand as ‘‘the Greek
city’’: these are of course the widespread private emigration to the provinces
by Roman citizens from Italy, and their settlement in Greek cities; and the
steadily increasing scale of the granting of Roman citizenship to individuals,
and hence to their descendants, in the Greek cities. It is not necessary here to
accumulate references to modern studies of these two processes; but it may
be noted simply that our knowledge of both, as regards the imperial period,
is very largely a function of the sudden explosion of the ‘‘epigraphic habit,’’
referred to above.
Taken all together, however, these processes, along with others to which
we will come, meant that ‘‘the Greek city’’ of the imperial period would be
more correctly described as ‘‘Graeco-Roman’’: that is, as a fusion or mélange
of languages and constitutions, types of public entertainment, architectural
forms, and religious institutions. The role of colonisation within this process of fusion has perhaps not been sufficiently stressed, so an example will
be given from a series of second-century inscriptions from the colony of
Cremna in Pisidia. In principle of course, colonies were supposed to use Latin
in their public life. An important Augustan inscription from the colony of
Alexandria Troas indeed shows this rule in operation. It honours C. Fabricius
Tuscus, duovir (one of the two annual magistrates) and augur (priest) in the
colony, who apart from a long list of military functions had been ‘‘prefect . . .
in charge of public works carried out in the colony by order of Augustus’’
( praef(ectus) . . . operum quae in colonia iussu Augusti facta sunt); the inscription
was put up d(ecreto) d(ecurionum) (by a decree of the senators).31
Elsewhere, however, and no doubt in Alexandria Troas too with the passage of time, Greek tended to reassert itself, while still being deployed to
express those new and distinctive institutions which had been created by the
Pisidia, Cremna, Olbasa, Comama, Lystra, Parlais. For the background, see S. Mitchell,
‘‘The Hellenization of Pisidia,’’ Mediterranean Archaeology  (): .
. Text to nn. – below.
. AE , no.  I. K. Alexandria Troas, no. .
The Greek City in the Roman Period

establishment of the colony. Hence for instance there is a group of inscriptions from Cremna, which come from a series of statue bases, of Herakles,
Nemesis, Athena, Hyg(i)eia, and Asklepios, put up by the colony.32 One example will suffice to illustrate the fusion of languages and concepts involved:
τὸν Ἡρακλέα
ἡ κολωνία
δυανδρία⟨ι⟩ς πενταετηρικῆ[ς]
τῶν ἀξιολογωτάτων Φλα.
Ἀουιδίου Φαβιανοῦ Καπιτωνιανοῦ Λουκίου καὶ Ῥοτειλιανοῦ Λογγιλλιανοῦ
Καλλίππου
The colony (honour) Herakles at the time of the quinquennial
duovirate / of the most worthy Flavius / Avidius Fabianus
Capito / nianus Lucius and of Ruti / lianus Longillianus /
Callippus.
The deities honoured are all addressed in purely Greek names. But otherwise
the names of the duoviri, Flavius Avidius Fabianus Capitolianus Lucius (?) and
Rutilianus Longillianus Callippus, are almost entirely Latin, using both the
lengthened -anus forms and the extended combination of names characteristic of the later imperial period. Colonia is duly transliterated as κολωνία.
More than a century and a half after the foundation of the colony, Cremna
could not be characterised either as ‘‘Greek’’ or as ‘‘Roman’’; for it was evidently both.
Colonisation was, however, a relatively isolated example of a positive
measure taken by Rome which had the immediate effect of introducing Roman, and Latin, elements into a Greek social and cultural environment. Paradoxically, it was far outweighed in its effects by the creation of new Greek
cities, both by emperors and by dependent kings; foundations by both often
had mixed, Graeco-Latin, names borrowed from those of the ruling imperial
dynasty. The first and most prominent of the imperial foundations was of
course Nicopolis, founded to commemorate Actium, created by the concentration of population, involving a re-distribution of votes in the Delphic
amphictyony (the group of cities with rights at the shrine), and giving rise to
a new and central element in the circuit of Greek athletic and theatrical fes. G. H. R. Horsley, ‘‘The Inscriptions from the So-called ‘Library’ at Cremna,’’ Anat.
Stud.  ():  SEG XXXVII, nos. –. The example given is no. ; see now
I. K. Central Pisidia, no. .

The Hellenistic World and Rome
tivals, the ‘‘Actia.’’ It seems to be as early as the s .. that we have the first
reflection of this new element, in the inscription from Pergamon recording the victories of Glycon at the Olympian, Pythian, Actian, and Nemean
games.33 But in spite of its name, and its new role among Greek cities, it can
be argued that Nicopolis was a double community, both a Greek city and a
Roman colony, not physically distinct.34
If it were really so, it was a rare case. More commonly, both emperors and
kings spread over the map of the Greek world an ever-denser network of
Greek cities, whose Greek names incorporated Latin, that is, (almost always)
imperial, personal names. It is superfluous to give more than examples, since
the process was detailed with remorseless care by A. H. M. Jones: Sebastē,
Caesarea, Tiberias, Germanicia, Claudiopolis, Flaviopolis, Flavia Neapolis,
Traianopolis, Marcianopolis, Hadrianopolis and Hadrianoutherae, and so on
to Philippolis and two places called Maximianopolis, a Diocletianopolis, and
of course Constantinopolis itself.35 The real nature of such re-foundations
cannot be pursued further here, for only very detailed local examination
would serve to make clear how far in each case urbanisation, in various forms,
had preceded the creation of the ‘‘new’’ city, and how radical the changes in
local social structures were. We cannot always distinguish a new foundation,
and the creation of a new urban structure, from the mere acquisition of a new
imperial name, as when Palmyra became ‘‘Hadrianē Palmyra’’; or be certain
as to how much new building accompanied any such transformation, or what
role was played by imperial initiative and benefaction.36 What is clear in two
well-known cases is that an imperial decision to grant the status of city to
an existing community could be a response to initiative from below. Hence
the letter, in Latin, addressed to a governor by an emperor whose name is
lost, agreeing that Tymandus in Pisidia has fulfilled the criteria for achieving
city status; in this case the availability of sufficient persons (fifty, initially) to
act as decuriones (members of a local senate), pass decrees, and elect magis. See E. Chrysos, ed., Nicopolis I: Proceedings of the First International Symposium on Nicopolis (). For the inscription from Pergamon, see L. Moretti, Iscrizioni agonistiche greche
(), no. .
. For this view see N. Purcell, ‘‘The Nicopolitan Synoicism and Roman Urban Policy,’’
in Chrysos (n. ), .
. A. H. M. Jones, The Cities of the Eastern Roman Provinces 2 ().
. See esp. S. Mitchell, ‘‘Imperial Building in the Eastern Roman Provinces,’’ HSCPh
 (): , with a shorter version in S. Macready and F. H. Thompson, eds., Roman Architecture in the Greek World (), , a valuable series of studies on the influence of Rome
on the physical character of imperial Greek cities.
The Greek City in the Roman Period

trates.37 Different criteria are set out by Constantine in allowing the claim
of Orcistus in Phrygia: its previous possession of the rank of civitas (a city
status), its location at the meeting place of four roads, a well-watered site—
and the fact that the population are Christian.38 Both imperial communications embody a general statement of favour towards the creation of new
cities (civitates/poleis), without allowing us to determine on whose initiative
such a creation normally depended.39
The question of construction, urban development, and architectural techniques will also not be pursued further here, except to emphasise that it was
not merely colonies which we would do better to characterise as ‘‘GraecoRoman’’ rather than as ‘‘Greek’’ cities. For even the ‘‘Greek’’ cities tended to
exhibit a range of ‘‘Graeco-Roman’’ features: theatres, usually of the ‘‘Roman’’ type with a raised stage; temples on a raised podium, with a frontal
axis; occasional amphitheatres; and more often perhaps theatres adapted to
accommodate gladiatorial shows or wild-beast hunts; baths and colonnaded
streets—all these represented a language of urbanism and public architecture which was shared with the generally less developed cities of the Latinspeaking West.40
Nor is it necessary to review here the administrative and political aspects
of the role of Greek cities within the Roman Empire, or their internal constitutional and financial structure. It may be sufficient to note that the independence of the last fully self-governing koinon (league of cities), that of
Lycia, was removed by Claudius in .. , when the area became part of the
new province of ‘‘Lycia et Pamphylia.’’ 41 The koinon itself continued, though
. MAMA IV, no. .
. MAMA VII, no. .
. See now the new inscription from Heraclea Sintica in Macedonia: G. Mitrev, ‘‘Civitas Heracleotarum: Heracleia Sintica or the Ancient Village of Rupite,’’ ZPE  ():
–, with C. Lepelley, ‘‘Une inscription d’Heraclea Sintica (Macédoine) récemment découverte, révélant un rescrit de l’empereur Galère restituant ses droits à la cité,’’ ZPE 
(): –.
. For the most penetrating study of urban architecture, drawing equally on Greek
East and Latin West, see W. L. MacDonald, The Architecture of the Roman Empire II: An Urban
Appraisal (). For the widespread adoption in the Greek East of what are categorised as
‘‘Roman’’ architectural techniques, see H. Dodge, ‘‘The Architectural Impact of Rome in the
East,’’ in M. Henig, ed., Architecture and Architectural Sculpture in the Roman Empire (), .
. Suetonius, Div. Claud. ; Dio , , –. The fullest account so far available is RE,
s.v. ‘‘Lykia,’’ supp. XIII (), cols. – (S. Jameson). Subsequent discoveries will allow
a major new account. For the formation of the province, see D. Magie, Roman Rule in Asia
Minor (), .

The Hellenistic World and Rome
with features which made it not entirely typical of the ‘‘provincial’’ koinon
familiar from other areas; nothing of significance can yet be added to the
study of these by J. Deininger.42
As regards individual poleis, it is beyond question that, while the formal
constitution found almost universally was that of magistrates (archontes),
council (boulē ), and people (dēmos), the council was normally now a body
whose members retained their position for life, and represented the upper
class of the community; it is also noticeable, as we have seen, that it was assumed that the prospective council (curia) of Tymandus would be responsible
for electing magistrates. A tendency towards oligarchic regimes determined
by class and wealth is thus undeniable, without its being demonstrable, except in the case of Pontus and Bithynia, that such a non-democratic regime
was actually instituted by Roman regulation.43 The question of for how long
the assembly of the people (ekklēsia) of a typical polis continued to meet, and
what real powers it will have exercised, would deserve further examination.
What is in any case certain about the Greek, or Graeco-Roman, city of the
imperial period is the central place occupied by the council (boulē ). That
essential role is mirrored by the vast range of archaeological evidence for city
council houses (bouleuteria or curiae), collected and assessed for the first time
by Jean-Charles Balty.44
In any case the broad themes of the political functioning of the Greek
cities, treated in the great work of A. H. M. Jones, have more recently been
re-examined, as regards the period between Augustus and Severus Alexander, in a masterly study by Maurice Sartre.45 It is particularly valuable that
Sartre has been able to place the ‘‘Greek city’’ of this period within the framework both of the geographical and administrative evolution of the Empire
on the one hand, including the role and progressive disappearance of ‘‘client
kingdoms,’’ and of a series of regional studies on the other. The areas treated
are Greece and Macedonia; Thrace and Moesia Inferior; Asia Minor; Syria
and Arabia; Judaea (and the Jewish diaspora); and Egypt. The Greek cities of
. J. Deininger, Die Provinziallandtage der römischen Kaiserzeit (). For the Lycian
koinon, ff.
. For a very careful collection of the evidence, see G. E. M. de Ste Croix, The Class
Struggle in the Ancient Greek World (), app. IV: ‘‘The Destruction of Greek Democracy
in the Roman Period.’’
. J.-M. Ch. Balty, Curia Ordinis: recherches d’architecture et d’urbanisme antiques sur les curies
provinciales du monde romain ().
. A. H. M. Jones, The Greek City from Alexander to Justinian (); M. Sartre, L’Orient
romain: provinces et sociétés provinciales en Méditerranée orientale d’Auguste aux Sévères ( avant
J.-C.– après J.-C.) ().
The Greek City in the Roman Period

the western Mediterranean and the Adriatic thus inevitably remain outside
his brief. So, more regrettably, do those of the Bosporan kingdom, which
have produced a substantial crop of inscriptions;46 those of Cyrene, whose
inscriptions have yet to be collected in a corpus, but which include important items, for instance of course the five ‘‘Cyrene edicts’’ of Augustus; and
letters of Hadrian and Antoninus Pius on membership of the Panhellenion,
and on the question of where jurisdiction should be given by the proconsul
of the geographically divided province of Crete and Cyrene.47
For obvious reasons the Greek cities of the Parthian Empire also lay outside the scope of Maurice Sartre’s work. The available evidence on Seleucia
on the Eulaeus (Susa), Spasinou Charax (Mesene), or the cities of Babylonia,
notably Seleucia on the Tigris, as they were in the first three centuries ..,
has not increased greatly in recent years.48 Nor has much greater attention
been paid to those places which in the second and third centuries were incorporated in the Roman Empire. The most notable is of course Dura-Europos,
the study of which, as it was both in the Parthian and the Roman periods, has
been gravely hampered by the failure to produce a corpus of its inscriptions,
in an unprecedented range of different languages.49
Somewhat more progress has been made with the Greek cities produced
by Hellenistic colonisation in northern Mesopotamia. Long part of the Parthian Empire, they came under Roman rule with the conquests of Septimius Severus in the s, and many then underwent a rapid transformation
into Roman colonies, producing documents in an inextricable mélange of
Greek, Syriac, and Latin.50 Like the ‘‘Greek cities’’ of other regions, they
were carefully treated in A. H. M. Jones’ great survey; but both their character as self-governing ‘‘Greek’’ cities and their bilingual Syriac/Greek culture
. V. V. Struve, Corpus Inscriptionum Regni Bosporani (CIRB) (); see V. F. Gajdukevi,
Des bosporanische Reich ().
. For the five Cyrene Edicts, see still the discussion by F. de Visscher, Les édits d’Auguste
découverts à Cyrène (). For the letters of Hadrian and Antoninus Pius, see J. M. Reynolds,
‘‘Hadrian, Antoninus Pius and the Cyrenaican Cities,’’ JRS  (): . The Jewish inscriptions are collected by G. Lüderitz, Corpus jüdischer Zeugnisse aus der Cyrenaika (). A
corpus of the inscriptions of Cyrenaica is being prepared by J. M. Reynolds.
. See still, for some of them, N. Pigulevskaya, Les villes de l’état iranien aux époques parthe
et sassanide (). See also S. A. Nodelman, ‘‘A Preliminary History of Characene,’’ Berytus
 (): .
. See F. Millar, ‘‘Dura-Europos under Parthian Rule,’’ in J. Wiesehöfer, ed., Das Partherreich und seine Zeugnisse (Historia-Einzelschrift , ), – ( chapter  of the present volume).
. For these places as coloniae, see F. Millar (n. ).

The Hellenistic World and Rome
render them mysterious, if extremely interesting, and their history will not
be pursued here.51 But it should be noted that the history of this region is
now illuminated by the publication of a new archive of Greek and Syriac
documents from the Middle Euphrates of the s to s.52
This rapid sketch is intended only as a reminder that, fundamental as
Maurice Sartre’s L’Orient romain is, its programme is dictated by the geographical shape of the Roman Empire up to ; if its subject had been not
‘‘the Roman Orient’’ but ‘‘the Greek city,’’ it would have been relevant to
explore some more marginal zones of the existence of the ‘‘Greek city’’; perhaps ‘‘the frontiers of Greek city life’’ might be a topic, another day, for a
different colloquium.
As it is, it is inevitable that if we try to characterise the most salient features of the Greek city of the Roman imperial period, we should focus on
the central zone, Greece (Achaea) itself, Macedonia, the province of Asia,
Pontus and Bithynia, Lycia, and to a lesser extent the other provinces of Asia
Minor. Our conceptions are necessarily dependent on the survival of literature which illuminates these places, the presence of substantial bodies of
inscriptions, and the existence of physical remains. In large and important
areas when substantial numbers of ‘‘Greek cities’’ were to be found, such as
central Asia Minor or the Near East west of the Euphrates (the Roman provinces of Syria, Judaea/Syria Palaestina, and Arabia), these conditions are only
partially present, and historical study is only now beginning to bring some
aspects of social and cultural history into focus.53
None the less, there is one medium by which the Greek cities of the Roman period, in all areas, expressed their identity, and whose products are
available to us in vast numbers, and independently of the types of evidence
mentioned above. This is of course the coinage of the cities, whose significance requires some emphasis. Ramsey MacMullen’s phrase ‘‘the epigraphic
habit’’ has justly established itself as a key element in our perception of the
Roman Empire and its cities.54 For the sheer scale of communal and individual self-expression through the medium of public writing in permanent
. See Jones (n. ), chap. ; F. Millar, The Roman Near East ( ..–.. ) (),
part. II, chap. .
. See D. Feissel and J. Gascou, ‘‘Documents d’archives romains inédits du Moyen
Euphrate (III e s. après J.-C.) I. Les petitions (P.Euphr.  à ),’’ JS : –; ‘‘II. Les actes
de ventes-achat (P.Euphr.  à ),’’ JS : –; ‘‘III. Actes divers et letters (P.Euphr.  à
),’’ JS : –.
. For central Asia Minor, note S. Mitchell, Anatolia: Land, Men and Gods in Central Asia
Minor,  ..–..  I–II (), and for the Near East, Millar (n. ).
. R. MacMullen, ‘‘The Epigraphic Habit in the Roman Empire,’’ AJPh  (): .
The Greek City in the Roman Period

form is one of the most striking features of imperial city culture. But so also is
the profusion of coinage, with the extra dimension that it involved the selection of visual images to accompany the (inevitably brief ) written legends.
The figures are remarkable: more than  Greek cities or city leagues (koina)
minted coins at one time or another, in the first three centuries of the Empire, and the largest total for a single reign was reached under Septimius
Severus (.. –), namely something over .55 The result is an extraordinarily rich repertoire of images-plus-legends deployed to express the
collective identity of cities; to portray and sometimes name the deities which
were important to them; and on occasion to commemorate recurrent events,
such as festivals, which played a central point in their collective existence, or
individual events, such as visits by emperors. It is important to stress that the
city coins, almost all only in bronze, were produced discontinuously; their
strictly economic function, and their role within the framework of properly
‘‘Roman’’ coinage, in gold, silver, and bronze, awaits serious analysis. More
important is the conclusion which has to be drawn from the late Konrad
Kraft’s study of the coinage of Asia Minor.56 He showed that the same workshop (however we might imagine the physical reality of a ‘‘workshop’’) might
produce coins for two or more different cities. We cannot therefore speak of
‘‘the mint’’ of Ephesos, or of any other city, but only of coins ‘‘of ’’ Ephesos; and
this term can be used legitimately only where the city (or rather community; see below) concerned is explicitly named. Deductions of the form that
(for instance) silver tetradrachms struck under Caracalla, but naming no city,
were produced by ‘‘the mint of Laodicea,’’ because the types resemble those
on coins of Laodicea, are wholly illegitimate.57 We have no basis for imagining the existence of stable city ‘‘mints’’; all that we can know is the coins
themselves, as extremely explicit and—in artistic terms—often very refined
and beautiful expressions of collective identities and values. How, where,
and by whom they were actually produced or manufactured is a matter of
speculation.
Until recently the best introduction to these coins, the ‘‘Greek Imperials,’’
has been in a well-illustrated study by K. Harl of the Greek city coinages of
the period from ..  to their complete disappearance in approximately
the third quarter of the third century.58 It is a pity that this fine study, con. T. B. Jones, ‘‘Greek Imperial Coins,’’ North American Journal of Numismatics  ():
; see A. Johnston, ‘‘Greek Imperial Statistics: A Commentary,’’ RN  (): .
. K. Kraft, Das System der kaiserzeitlichen Münzprägung in Kleinasien ().
. For these misleading presumptions, see, e.g., A. Bellinger, The Syrian Tetradrachms of
Caracalla and Macrinus ().
. K. Harl, Civic Coins and Civic Politics in the Roman East, .. – ().

The Hellenistic World and Rome
sistently trying to set this coinage in a wider social and ideological context,
does not attempt to explain the reasons for its disappearance (other than the
depreciation of the imperial silver coinage) or the economic effects of the
cessation of local minting. But for the latter question it would be necessary
to distinguish minting—the production of new coins—from circulation—the
use as a medium of exchange of coins of various origins and dates.
If the economic effects of this great change are highly uncertain, the abandonment of local coining cannot fail to be seen as the loss of a crucial means
of self-expression by several hundred Greek cities. In the longer term, it
can be taken as an aspect of the transformation of later Greek cities from
pagan communities, symbolised above all by images of their deities and of the
temples which housed their cult statues, into Christian communities under
bishops. But since city coining was never to revive, no Greek city in the Roman Empire ever had occasion to adorn its coins with Christian images or
legends.
Our knowledge of city minting in the early part of the period, from circa
 .. to .. , has been transformed by the publication of the first two
parts of a really major project, The Roman Provincial Coinage.59 For the first
time all the provincial coinages of the first century of the Empire have been
assessed and catalogued; further volumes (how many?) will carry the story
to the cessation of minting in the later third century.
It is too early to assess in any detail the significance of this already gigantic
contribution, which at a stroke transforms our ability to envisage the Roman
Empire from the standpoint of hundreds of separate provincial communities. But three features stand out. Firstly, the scale and geographical range of
coining. In the first two-thirds of a century of the Empire local communities
in Spain and Africa might also produce coins. But by the reign of Claudius
(.. –) they fall silent, and coinage becomes one of a range of ways in
which the Roman Empire gave, or allowed, a specially privileged status to
Greek cities. Beyond that, the number of Greek cities producing coins increased rapidly in the third century, to reach (as mentioned above) a maximum of more than  under Septimius Severus. At that moment it was at
least one-and-a-half centuries since any community which was not Greek,
and which belonged to the Roman Empire, had minted its own coins. The
exception proving the rule is of course the Hebrew coinage of the Jewish
revolt and the Bar Kochba war; but at those moments, each prolonged for
several years, the community concerned did not ‘‘belong’’ to the Empire.
. A. Burnett, M. Amandry, and P. P. Ripollès, The Roman Provincial Coinage I.––
II (–).
The Greek City in the Roman Period

The second feature which is (almost) all-pervasive is the name and image
of the reigning emperor. Except where local eras were used, it is of course
precisely the presence of this name and image which allows us to arrange
the coins of any one community in a temporal sequence. The naming and
portrayal of the emperor is not universal: there is a large category of city
coins from all areas of the Greek East in which other images (most often
of gods) and legends are substituted for the portrait and name of the emperor. Such coins are conventionally labelled ‘‘pseudo-autonomous,’’ on the
presupposition that they embody some special privilege, or special degree of
freedom from Roman control: but none such can in fact be identified. None
the less, it is surely significant that Tyre is both the only ‘‘Greek city’’ in the
Empire which continued to use some non-Greek (Phoenician) letters on its
coins right up to the moment of its transformation in the s into a Roman
colony; and that it was one of only three cities (the others being Chios and
Athens) which never (up to the same point) named or portrayed the reigning
emperor on its coins.60
The presence of the name and the image of the emperor has to be taken as
one of the dominating features of the collective life of the Greek city in the
imperial period. This applies, as noted above, very widely to the names of
the cities themselves, and not merely to those transformed into Roman colonies; to the personal names of individual citizens, in which the family names
(nomina) of successive ruling houses—‘‘Iulius,’’ ‘‘Claudius,’’ ‘‘Flavius,’’ ‘‘Ulpius,’’
‘‘Aelius,’’ ‘‘Aurelius,’’ ‘‘Septimius’’—were ever more prominent, transliterated
into their Greek forms; to the cults and temples of the emperors, reigning
or deceased, and individual or collective (‘‘the Sebastoi’’); to the identities of
public buildings, like the ‘‘Hadrianic Baths’’ revealed at Antioch in Syria by
a new document of .. ;61 to the names of months in city calendars; to
the names of tribes or other sub-units of the communities; to the names of
festivals; to the actual clothing of agōnothētai (president of games) or archiereis
(high priests; see below); to the presence of honorific statues of emperors
and members of their families; and to the prominence of inscribed letters
from emperors, written in Greek, and of other inscriptions recording privileges granted to individuals by emperors. It is not too much to say that the
public self-expression of the ‘‘Greek city’’ in the Empire embodied at every
level an explicit recognition of the distant presence of the emperor. When
. See A. Johnston, ‘‘The So-Called ‘Pseudo-Autonomous’ Greek Imperials,’’ Am. Num.
Soc., Mus. Notes  (): . There is no complete study of the coinage of Tyre; for the
essentials, see BMC Phoenicia cxxiii–cxxiv, and –; for these three cases, see RPC I, .
. See Feissel and Gascou (n. ) I (), no. .

The Hellenistic World and Rome
the emperor travelled, that presence might become real; but our evidence,
however biased, is enough to show that a real, concrete connection between
city and emperors was maintained far more intensively by the constant traffic in embassies, whether on diplomatic missions or in pursuit of requests or
disputes, to bring city decrees before the emperor, address him in person,
and bring back a letter in Greek in reply.62 As we see in the case of a letter
of Caracalla of .. , even an imperial letter to an individual, referring
to the obligations to his patris (home city) on the part of another individual,
might be ‘‘read out in the theatre,’’ in this case that of Philadelphia in Asia.63
The purpose of doing that was of course to publicise it before the citizens,
meeting in the theatre, ‘‘where it is their custom to take counsel,’’ as Tacitus
wrote of the Antiochenes in Syria.64 When aroused, the citizens might even
gather spontaneously in the theatre, as Acts represents the Ephesioi doing
when provoked by Paul’s teaching.65 But these allusions reflect a much more
important truth about what we call, in some ways misleadingly, a ‘‘Greek
city.’’ A Greek city was not essentially a ‘‘place’’ or an urban centre; it was a
community of individuals. The point is made with great clarity, but without
further development, in an important chapter by Joyce Reynolds on ‘‘Cities’’
in the context of the administration of the Empire.66 Thus when an emperor
wrote, as we would say, ‘‘to Pergamon,’’ that was not in fact how he expressed
himself: his letters would go to ‘‘the archontes, boulē, and dēmos of the Pergamēnoi’’ (to the magistrates, council, and people of the Pergamenians); or,
if ‘‘to Aphrodisias,’’ ‘‘to the archontes, boulē, and dēmos of the Aphrodisieis’’ (to
the magistrates, council, and people of the Aphrodisians). The point is not a
trivial one, for we consistently mistranslate, and therefore misconceive, the
nature of the communal attachments which gave people their identity, in the
. This traffic in embassies is a central theme of my The Emperor in the Roman World,
 ..–..  (; reissued with afterword, ). For imperial letters, see the posthumously published corpus by J. H. Oliver, Greek Constitutions of the Early Roman Emperors from
Inscriptions and Papyri ().
. IGR I, no. ; Dittenberger, Syll.3, no. ; Abbott and Johnson, Municipal Administration, no. .
. Tacitus, Hist. , .
. Acts :. It is worth noting the valuable survey by G. H. R. Horsley, ‘‘The Inscriptions of Ephesos and the New Testament,’’ Novum Testamentum  (): .
. J. Reynolds, ‘‘Cities,’’ in D. C. Braund, ed., The Administration of the Roman Empire
(Exeter Studies in History , ), : ‘‘They were, strictly speaking, ‘peoples’ ( populi,
demoi ), properly designated by an ethnic rather than a place name (Carthaginienses rather
than Carthago, Pergamenoi rather than Pergamon).’’
The Greek City in the Roman Period

eyes of both themselves and others. Confronted with ‘‘Dionysios the Halikarnasseus,’’ we invariably write ‘‘Dionysius of Halicarnassus,’’ as if what he
belonged to were a point on the map, but it was not; it was a community of
citizens.
This point is consistently reinforced by the extremely extensive evidence
of communal designations on the coinage ‘‘of cities.’’ Here too, we would
be less liable to confuse ourselves if we called it the coinage of ‘‘communities.’’ The male genitive plural is by far the most common grammatical form
of identification on the coins: so, to take only some cases from the index of
Roman Provincial Coinage I ‘‘of the Aizanitai,’’ ‘‘of the Alabandeis,’’ ‘‘of the Amisoi.’’ There are exceptions, though they are far less common; occasionally we
do find a place-name used to identify the origin of a coin. So, for example,
from the same source ‘‘Amisos,’’ ‘‘Gadara,’’ ‘‘Gerasa,’’ ‘‘Thessalonikē.’’ But the
fact that we now can, at least for the first hundred years of the Empire, read
right across the entire local (or communal) numismatic production of the
provinces is the first great service which Roman Provincial Coinage I rendered.
In a profound sense, an important means of approach to the communities of
the Empire, and to the ‘‘Greek city’’ above all, has been opened up for the
first time.
I have placed a lot of emphasis on this material, precisely because in this
now organised and intelligible form, it is new. But the Greek city of the Empire has of course revealed itself to us primarily via inscriptions. It is impossible to sum up the wealth of information contained in the tens of thousands
of Greek city inscriptions of the imperial period, which now constitute a
genre of literary expression in themselves. Instead, it may be preferable to
attempt to present, and bring out the significance of, a few choice examples.
The two most revealing both illustrate the mass, collective character of the
life of the Greek city (or community), a point heavily and correctly stressed
by Ramsey MacMullen and Robin Lane Fox in their studies on the paganism
of this period.67 But they also reflect the way in which the collective ceremonials and observances of Greek communities under the Empire gave a special
place to the figure of the emperor.
The first inscription comes from a relatively little-known corner of the
Greek world, the modest city of Kalindoia in Macedonia.68 With a certain
. R. MacMullen, Paganism in the Roman Empire (); R. Lane Fox, Pagans and Christians
(), esp. –, ‘‘Pagans and Their Cities.’’
. Edited by K. L. Sismanides in Ἀρχαιολογικόν δελτίον Μελέτες  []: ; see
BE : no. , and SEG XXXV, no. .

The Hellenistic World and Rome
poetic appropriateness, it dates from the first year of the Christian era, and
thus forms a pair with a comparable inscription from Kyme, of  ..–.. .69
Like thousands of other inscriptions of the period, the latter honours a euergetēs (benefactor) who had offered hospitality and public shows to the people;
moreover ‘‘during the Kaisareia [a festival in the name of the emperor] celebrated by Asia, as he had promised, he carried out sacrifices and banquets
where the flesh of the victims was consumed, having made a sacrifice of oxen
to the emperor Kaisar Sebastos and to his sons and to the other gods.’’
But this document, significant only for being so characteristic of the epigraphic expression of the next three centuries, is far exceeded in importance
by that from Kalindoia, which presents in concentrated form, and very early,
almost all the values of Greek communal life, in its relation to the Empire.
It therefore deserves translation in full:
Year  [of the provincial era of Macedonia, .. ]
The city magistrates [ politarchai], after a preliminary resolution by the
members of the council [bouleutai], and an assembly of the people [ekklēsia] being held, declared before the people [dēmos]:
Since Apollonios son of Apollonios son of Kertimos, being a good
man and deserving of every honour, having accepted spontaneously
the priesthood of Zeus and Rōmē and Caesar Augustus divi filius (son
of the deified [Iulius]), has exhibited so much nobility, living up to the
high reputation of his ancestors and of his own virtue, as to omit no
excess of expenditure on the gods and his native city, providing from
his own resources throughout the year the sacrifices offered monthly
by the city to Zeus and Caesar Augustus; and has also offered all manner
of honours to the gods, and provided for the citizens feasting and lavish
entertainment, similarly dining the whole populace, both en masse and
by triklinia (separate dining groups), and organising the procession at
the festival so as to be varied and striking, and putting on the contests in
honour of Zeus and Caesar Augustus in elaborate and worthy style . . .
has shown his generosity to his fellow citizens by asking from the city
leave to take over the public sacrifices offered during the festival to
Zeus, Caesar Augustus, and the other benefactors, and has provided
them at his own expense; and having sacrificed oxen has entertained
. R. Hodot, ‘‘Décret de Cymè en l’honneur du Prytane Kléanax,’’ J. Paul Getty Museum Journal  (): , with the long discussion by L. Robert, Bull. Épig. , ;
R. Merkelbach, ‘‘Ehrenbeschluss der Kymäer für den Prytanis Kleanax,’’ Epigr. Anat.  ():
; SEG XXXII, no. .
The Greek City in the Roman Period

each of the citizens throughout the whole festival, by triklinia and on a
mass basis, and made the most lavish distributions to the tribes, so that,
wherever they wished to take their pleasure, they did so by his grace.
Not only has he spared no expense, but he has had a statue of Caesar
made at his own cost, and has offered it as a permanent memorial of
the beneficence of Augustus to all mankind; he has thus provided an
additional ornament for his native city, and for the god the appropriate
honour and favour.
For these reasons it seems appropriate to the council and people to
praise him for the enlightenment of his spirit and of his generosity
towards his native city, to crown him with a wreath and to vote a stone
(marble?) image of himself, of his father Apollonios and his mother
Stratto; to set up the statues and the decree in whatever place in the
agora the agōnothetēs [Apollonios] chooses, in order that other citizens
might be rendered eager to seek honour and to contribute generously
to their native city.
When the decree was voted (by the assembly), Apollonios accepted
the honour and the gratitude of his homeland, but relieved the city of
the expense.
It was voted on the th of Daisios.
This very early inscription encapsulates almost all the key features of the
public life of the imperial Greek city: the role of festivals and of public, communal celebrations; the importance of public writing and publicly placed
images; the ‘‘presence’’ of the absent emperor, both as an object of worship
and as visibly represented in his statue; and the central significance of the
complex symbolic, political, and financial exchanges between leading individuals and the mass of their fellow citizens which made up the institution
of euergetism. Given the early date, one element is missing: the progressive
extension to the leading families in most Greek cities of the Roman citizenship, and, following on that the acquisition by their members of positions in the equestrian service and the Roman Senate. This steady evolution,
known to us almost entirely through honorific inscriptions put up by their
native cities, is familiar, and need not be re-examined here.70 The effect was,
. A detailed study of the equestrian roles of men from the Greek cities is lacking.
See the classic sketch by A. Stein, ‘‘Zur sozialen Stellung der provinzialen Oberpriester,’’
Epitumbion Swoboda (), ; note also F. Quass, ‘‘Zur politischen Tätigkeit der munizipalen Aristokratie des griechischen Ostens in der Kaiserzeit,’’ Historia  (): , esp.
–, for holders of equestrian posts. For senators, H. Halfmann, Die Senatoren aus dem
östlichen Teil des Imperium Romanum bis zum Ende des z. Jahrhunderts n. Chr. (); S. Pan-

The Hellenistic World and Rome
firstly, that in all cities the ruling circle of office-holders and members of the
council showed a steadily increasing proportion of Latin names, normally in
the standard triple Roman form of praenomen, nomen, and cognomen (the latter
very frequently still a Greek name); and with that, in principle at least, the
application to their family structures and property relations of Roman private law (for instance, the father’s legal authority over his extended family
known as the patria potestas). But neither the rules of the pre-existing private law of the Greek cities nor the real extent of the currency of Roman
private law can be easily understood.71 The spread of the Roman citizenship
to individuals, and hence families, culminating in the Constitutio Antoniniana
(the universal grant of citizenship made by Caracalla in ), was however
yet another respect in which the ‘‘Greek’’ city of the imperial period was
in reality ‘‘Graeco-Roman.’’ A secondary effect was that those persons who
served as equestrians or senators ‘‘belonged’’ not only to their native cities,
but also to a much wider world, whose varied regions might be represented
before their fellow citizens in the record of the places where they had served,
incorporated in inscriptions put up in their honour.
The processes touched on above might have led to a wholesale loss, or
even suppression, of the historical identities of Greek cities, and their absorption into an undifferentiated Graeco-Roman culture and outlook; or on
the contrary, to a clear ideological ‘‘resistance,’’ and to a reassertion of historic
identity; to actual armed resistance on the part of cities or regions; or, perhaps, to a claim to Greek dominance within that Graeco-Roman world on
which the Empire had imposed so high a degree of unity. As regards the last
possibility, we have seen above that it was indeed the case that Greek cities
enjoyed privileges, like coinage, or the fact that emperors and governors paid
them the compliment of addressing them systematically in Greek, which did
indeed mark them out from the urban communities of the West. But their
status was thereby explicitly recognised, and no Greek cities, or groups of
cities, offered any parallel to the major Jewish revolts of .. – and –
. It was however not merely a strategic shift towards the eastern frontier
which meant that Constantine’s new capital would be situated in the Greek
world, on the route between the Danube and Euphrates, and would have a
ciera, ed., Epigrafia e ordine senatorio II (), –. See above all, however, for the local
activity of senators, W. Eck, ‘‘Die Präsenz senatorischer Familien in der Städten des Imperium Romanum bis zum späten . Jahrhundert,’’ in W. Eck, H. Galsterer, and H. Wolff,
eds., Studien zur antiken Sozialgeschichte: Festschrift F. Vittinghoff (), .
. For a sketch of the problem, see H. Galsterer, ‘‘Roman Law in the Provinces: Some
Problems of Transmission,’’ in M. Crawford, ed., L’impero romano e le strutture economiche e
sociali delle province (), .
The Greek City in the Roman Period

Greek name. It was also a symbol of the fact that in the end Graecia capta did
indeed imprison her captor.
As regards the first possibility, the signs of any real loss of cultural identity were indeed very few. Latin literature, for instance, seems to have gained
extraordinarily little currency in the Greek East; and there is no certain evidence even of the translation of Virgil until Constantine included part of a
Greek version of the Fourth Eclogue in his Address to the Assembly of the Saints.72
Roman law, however, as taught in the schools of the Roman colony of Berytus, did, from the third century onwards, act as a magnet for the youth of the
Greek cities.73 Looking in the reverse direction, it is of real significance that
the history of Rome is largely ‘‘constituted’’ for us by Greek writers of the imperial period, Dionysius from Halicarnassus, Plutarch from Chaeronea, Appian from Alexandria, and Cassius Dio from Nicaea.74 The latter three were
all Roman citizens, the middle two of equestrian rank and the last-named a
senator and consul.
None the less, the most interesting and significant new items of evidence,
or new studies of already-known evidence, of the last few years have tended
to show a reassertion of historic and mythological identity in the framework of the collective life of the Greek city. But this cannot properly be
seen as a movement ‘‘against’’ Rome (and here again the contrast with the
two great Jewish revolts is fundamental). The most systematic of all these reassertions, indeed, the Panhellenion, was the work of the emperor Hadrian.75
In more specifically local contexts, the symbolic and communal assertion
of a city’s identity will have been the work of its local governing class—
but in the cases known to us this was a class in which the Roman citizenship, and even Roman office-holding, were already widespread. At Athens
indeed, there are clear indications that actual Roman office was not sought
as frequently as elsewhere. None the less, the Roman citizenship was common among the upper classes, and the resistance to the Herulian invasion
of the s was led by a member of a family which had long held the citizenship, Herennius Dexippus.76 Athens remained, even after the Herulian
. Constantine, Oratio ad Coetum Sanctorum –.
. See Millar (n. ), –.
. E. Gabba, Dionysius and the History of Archaic Rome (); C. P. Jones, Plutarch and
Rome (); F. Millar, A Study of Cassius Dio (). See in general E. Gabba, ‘‘The Historians
and Augustus,’’ in Millar and Segal (n. ), .
. See A. J. Spawforth and S. Walker, ‘‘The World of the Panhellenion I. Athens and
Eleusis,’’ JRS  (): ; ‘‘II. Three Dorian Cities,’’ JRS  (): .
. See G. M. Woloch, Roman Citizenship and the Athenian Elite, .. – (); F. Millar (n. ), .

The Hellenistic World and Rome
invasions, a major Hellenic centre. But the most systematic communal recreation, or re-enactment, of the past which modern study has revealed was
that at Sparta. Spawforth’s study has shown, however, that this re-creation,
whose salient features drew tourists from all over the Greek world, did not
really reproduce at all precisely the institutions of archaic and classical Sparta,
and was the work, as elsewhere, of a local elite among whom the Roman citizenship was increasingly common. Note for example the inscription of the
early third century in which a group of his synarchontes (fellow magistrates),
all with Roman names, honour ‘‘Pop(lios) Mem(mios) Pratolaos also called
Aristokles Damarous, aristopoleiteutēs [best citizen], for his protection of the
Lykourgan customs, and his benevolence towards them.’’ 77
An equally vivid picture of the communal evocation of tradition is provided by Guy M. Rogers’ analysis of the foundation of Vibius Salutaris in
early second-century Ephesos.78 Salutaris was of course a Roman citizen and
equestrian office-holder; the route of the procession which his foundation
instituted would take the participants down streets which had been completely transformed by monumental public building in the course of the first
century ..; the statues to be carried included representations of Trajan and
Plotina, as well as the personified Senate, equestrian order, and the Roman
people; and the terms of the foundation were approved by the proconsul and
his legate. But the prime honorand was the main deity of the city, Artemis;
and the other statues included one of Enonymos, son of Kephios/Ouranos
and Gē; Pion; probably Androklos, the mythical founder of the city; and
certainly Lysimachos, the early Hellenistic re-founder, as well as Augustus
himself. The rituals celebrated the present as the fulfilment of a long history,
not an opposition between Greek past and Roman present.
But of course the primary place among all recent work on the Greek
city of the Roman period must belong to Michael Wörrle’s publication of
the foundation inscription of a new agōn (contest), the Dēmostheneia, set up
at Oenoanda under Hadrian by a leading local notable, and Roman citizen,
C. Iulius Demosthenes.79 This almost perfectly preserved inscription of 
lines perhaps surpasses all others in its expression of the structure and values
of the imperial Greek city. I will isolate only a few features. First of all, there
is something of which our evidence rarely gives so vivid an impression, the
. IG IV., no. . See Spawforth in P. Cartledge and A. J. Spawforth, Hellenistic and
Roman Sparta: A Tale of Two Cities (), –.
. G. M. Rogers, The Sacred Identity of Ephesos: Foundation Myths of a Roman City ().
. M. Wörrle, Stadt und Fest im kaiserzeitlichen Kleinasien: Studien zu einer agonistischen
Stiftung aus Oinoanda ().
The Greek City in the Roman Period

list of villages (kōmai), with associated monagriai (farmsteads?) in the territory
of Oenoanda which are listed as being due to contribute cattle for sacrifice
(lines –). Then there is the detailed specification of the content of the
various forms of competition, and the prizes to be attached to them, listed
in the chronological sequence which is to be followed (lines –):
.
.
.
.
.
.
.
.
.
.
.
Trumpeters and heralds.  den(arii).
Composers of Prose Encomia.  den(arii).
Poets.  den(arii).
Oboists. st prize  den(arii)., nd .
Comic Poets. st prize  den(arii)., nd .
Tragic Poets. st prize  den(arii)., nd .
Kitharodes. st prize , nd .
Open competition. st prize  den(arii)., nd , rd .
Mime-artists and acts and displays. No prizes.
Other acts giving pleasure to the city.  den(arii). in all.
Gymnastic competitions for citizens.  den(arii). in all.
The detailed specification given here makes possible, not to say imperative, a
comprehensive new study of the competitive festival in the Greek world of
the imperial period, in which the primacy continued to the end to be held
by the ancient games of Delphi, Olympia, Nemea, and the Isthmus, together
with the Actia of Nicopolis. There have been important preliminary studies:
for instance the excellent collection of agonistic inscriptions published by
L. Moretti in ; the two suggestive papers by the great Louis Robert, cited
by Stephen Mitchell in his review article on Wörrle’s book (with an English translation of the text); and the collection of the inscriptions relating
to the artists of Dionysus in the new edition of Pickard-Cambridge on the
dramatic festivals of Athens.80 But none comes anywhere near being the fullscale study of the hierarchy of types of festival (from the major ones to the
most local), as well as their geographical distribution and their allocation
over the calendar, which could now be undertaken. Their geographical distribution alone would almost serve to define the world of the Greek city:
from Zeugma (but not across the Euphrates) to Bostra (the Aktia Dousaria,
. L. Moretti, Iscrizioni agonistiche greche (); L. Robert, ‘‘Deux concours grecs à
Rome,’’ CRAI (): ; L. Robert, ‘‘Discours d’ouverture,’’ VIIIth International Congress of
Greek and Latin Epigraphy  (), ; S. Mitchell, ‘‘Festivals, Games and Civic Life in
Roman Asia Minor,’’ JRS  (): ; A. W. Pickard-Cambridge, The Dramatic Festivals
of Athens 2, revised by J. Gould and D. M. Lewis (; reissued with corrections and supp.,
), – (the inscriptions relating to artists of Dionysus).

The Hellenistic World and Rome
the festival for the Nabataean God, Dousara) Gerasa, and Tyre; in Cappadocia and all of Asia Minor; in Egypt (but only at a modest local level); at
Puteoli, Naples, and (from Domitian onwards, with the Greek-style Capitoline games which he founded) Rome itself; at Massilia (but also at any other
western Greek cities?).
Their distribution in space was also, at least as regards the more important contests, an allocation in time, as two inscriptions from Aphrodisias explicitly state. One of them, concerned with the setting-up of a festival in a
manner quite close to that at Oenoanda, seems to schedule the festival ‘‘before the [departure of the competitors] to Rome.’’ In another, a high priest
of Asia also makes regulations for a festival and lays down that it will be held
‘‘in the period between the celebration of the Barbilleia at Ephesus and [? the
provincial games] of Asia.’’ 81
The games set up at Oenoanda, like so many others, were due to the philotimia (love of fame) of a local notable who was also a Roman citizen; the
procedures for ratifying it involved other locals who were also Roman citizens; the terms were approved both by the governor and by Hadrian himself,
whose letter opens the inscription; and images of the Emperor were to be
deployed in the proceedings.
But perhaps the most telling detail in the great inscription from Oenoanda
relates to the golden crown which the agōnothetēs was to wear, with ‘‘relief portraits of the emperor Nerva Trajan Hadrian Caesar Augustus and our
leader, the ancestral god Apollo.’’ 82 This passing allusion has much greater
significance than might appear at first sight. The crown combining images
of a deity and the emperor, to be worn by an agōnothetēs at a festival, might
be taken as a symbol of that whole ‘‘Romano-Greek’’ complex of beliefs,
customs, and communal observances which constituted the collective life
of the ‘‘Greek’’ city in the imperial period. In archaeological terms it is also
noteworthy that comparable crowns are worn by some of the local notables
portrayed in the remarkable statuary from Aphrodisias of this period.83 But,
more significantly, a crown of precisely this type appears in one of the most
brilliant of literary evocations of the challenge which Christianity offered to
the beliefs, customs, and collective values of the Greek city, the Acts of Paul
and Thecla, of which one version at least was already in circulation by the
end of the second century. When Paul and Thecla reach Antioch (apparently
. Reynolds, Aphrodisias and Rome (n. ), no. , l. ; no. , ll. –.
. Lines – (trans. Mitchell, above, n. ).
. For a photograph and brief account of one of them, see K. T. Erim, Aphrodisias: City
of Venus Aphrodite (), .
The Greek City in the Roman Period

the one in Pisidia), Alexander, ‘‘one of the first of the Antiochenes,’’ lays hold
of her. But she resists, tearing his cloak and dashing from his head a golden
crown with an image of the emperor.84
The Acts of Paul and Thecla represents only one of a series of Christian portrayals of the life of the Greek city, of the crisis caused in it by the preaching
of Christianity, and of the profound transformations which then came about.
One of the most striking is the Life of Gregorius Thaumaturgus by Gregory
of Nyssa. Written in the s, this novelistic portrayal of Christian preaching in the mid-third century vividly evokes the public buildings and popular
festivals of cities in Pontus, and seems to be unique in explicitly portraying
the establishment of a calendar of Christian martyr festivals as a conscious
device to create a new ‘‘rhythm’’ of city life (the term used is very significant: metarrythmizōn).85 That the communal structures and the pagan calendar, or ‘‘rhythm,’’ of the Graeco-Roman city did in fact, over a period of not
much more than a century, succumb to the new Christian world of bishops,
churches, shrines of martyrs, and a new calendar of feasts is of course certain.
Julian’s short-lived attempt to revive it met with only a limited response. His
last letter, written as his army marched east from Antioch in , might be
taken as the epitaph of the pagan city:
From Litarbae I proceeded to Beroea, and there Zeus by showing a
manifest sign from heaven declared all things to be auspicious. I stayed
there for a day and saw the Acropolis and sacrificed to Zeus in imperial
fashion a white bull. Also I conversed briefly with the senate about the
worship of the gods. But though they all applauded my arguments very
few were converted by them, and these few were men who even before I spoke seemed to me to hold sound views. But they were cautious
and would not strip off and lay aside their modest reserve, as though
afraid of too frank speech.86
Looking at the pagan Graeco-Roman city of the fourth century, we might,
as Oswyn Murray has suggested to me, wish to emphasise the extraordinary
stability, or ossification, of culture and values which bound it to the classical
Greek city of some seven centuries earlier. On this view it succumbed because it did not change, and could not. On the other hand, we might rather
. Acts of Paul and Thecla – (Syriac text) quoted from Price (n. ), .
. Gregory of Nyssa, Vita Gregorii Thaumaturgi, PG XLVI, cols. –. For μεταρρυθμίζων, see col. . See R. Van Dam, ‘‘Hagiography and History: The Life of Gregory
Thaumaturgus,’’ Classical Antiquity  (): .
. Julian, Ep. , Loeb trans.; Bidez-Cumont, Ep. .

The Hellenistic World and Rome
emphasise its vast progressive diffusion since then, with the effect that the
Greek-speaking city now provided the primary form of identity for perhaps
 million people; the growth in size, architectural adornment, and urban
facilities, such as aqueducts, characteristic of very many of them; the wider
unity symbolised by the cycle of athletic and ‘‘musical’’ festivals; and their
involvement, in many different ways, practical and symbolic, in the Roman
Empire. It is no accident that the ‘‘Greek city’’ whose ruins we can still see
was the ‘‘Graeco-Roman city’’ of the imperial period. But Christianity was
to triumph all the same.
The preaching of Christianity was not of course the only crisis which
steadily transformed the Greek city in the later imperial period. Invasions
touched much of the Balkans, Macedonia, Greece, Asia Minor, and Syria;
the depreciation of the imperial coinage seems (see above) to have been the
factor which brought about the ending of city coinage; the sub-division of
provinces tended to bring the governors closer to the individual city, and it
is in any case noticeable that building in the cities of the later Empire depended more on the initiative of governors than it did on local benefactors.
Equally, the tradition of euergetism seems to have been profoundly damaged
by a change which the emperors themselves introduced, as a way of rewarding those who served under them. That is to say that from around .. ,
various civilian and military ranks in the imperial service started to be conceived of as conferring a permanent named status on their holders; and these
statuses in their turn came to confer a life-long immunity from magistracies (archai), liturgies (leitourgiai), and membership of the council (boulē ) in a
man’s city. The ‘‘flight of the councillors’’ into imperial office, so characteristic of the fourth century, was in fact an artificial creation by the emperors
themselves, whose consequences they tried in vain to limit.87
After the conversion of Constantine and his subsequent defeat of Licinius, the new freedom and imperial backing given to Christian communities could be followed, at first slowly and erratically, by Christian attacks on temples, leading sometimes to their destruction and replacement by
churches.88 None of this meant of course that what we call ‘‘the Greek city’’
simply vanished. On the contrary, work at, for instance, Athens, at Aphrodisias in Caria, and at Scythopolis in Syria Palaestina has demonstrated how
. These points briefly summarise the conclusions of F. Millar, ‘‘Empire and City, Augustus to Julian: Obligations, Excuses and Status,’’ JRS  ():  ( chapter  of F. Millar,
Rome, the Greek World, and the East II: Government, Society, and Culture in the Roman Empire).
. See G. Fowden, ‘‘Bishops and Temples in the Eastern Roman Empire, .. –,’’
JThSt  (): .
The Greek City in the Roman Period

complex a story of communal life, and even of extensive new construction
in the fourth to sixth and seventh centuries, remains to be written.89 In the
Syrian region, however, major changes in the pattern of urbanism may have
been under way even before the Islamic conquests of the seventh century;90
in Asia Minor, as the work of Clive Foss has suggested, the Persian invasions of the early seventh century may have marked a decisive change.91 But
these are as yet mere pointers. It remains the case that the ‘‘Greek city’’ which
we can know best is the Romanised Greek city of the high imperial period,
whose physical remains, inscriptions, and coins offer us a uniquely rich public ‘‘image,’’ which still leaves fundamental questions of social and economic
history unanswered. But a very different history, whose framework would be
a changed relationship to the now Christian emperors, could now be written, on the basis of changing types of evidence, to do justice to the complex
evolution of the ‘‘Greek,’’ or ‘‘Graeco-Roman,’’ city of the fourth to seventh
centuries.92
. For late Roman Athens, see esp. A. Frantz, The Athenian Agora XXIV: Late Antiquity,
.. – (); for Aphrodisias, see esp. C. Roueché, Aphrodisias in Late Antiquity (),
and Performers and Partisans in Aphrodisias in the Roman and Late Roman Periods (). For the
excavations at Scythopolis, see esp. Excavations and Surveys in Israel – (–): –.
Note () the mosaic inscription of building work carried out under the governorship of
Palladius Porphyrius, probably in the fourth century.
. See H. Kennedy, ‘‘From Polis to Madina: Urban Change in Late Antique and Early
Islamic Syria,’’ Past and Present  (): .
. C. Foss, ‘‘The Persians in Asia Minor and the End of Antiquity,’’ EHR  ():
, reprinted along with other papers in his History and Archaeology of Byzantine Asia Minor
(), no. I. Note also his studies of individual cities: Byzantine and Turkish Sardis ();
and Ephesus after Antiquity ().
. See for instance M. Whittow, ‘‘Ruling the Late Roman and Early Byzantine City:
A Continuous History,’’ Past and Present  (): .
This page intentionally left blank
PA RT I I
Rome and the East
This page intentionally left blank
 
Reflections on the Trials of Jesus *
Introduction
If anything at all is certain about the earthly life of Jesus, it is that he was a Jew
who expressed original and disturbing conceptions of what Judaism ought
to mean, and was executed on the orders of a Roman praefectus who had little
or no conception of what Judaism meant. The varied and contradictory accounts which the Gospels provide of how Jesus came to suffer crucifixion
may thus be a suitable topic for me, as a Roman historian, to offer in honour of Geza Vermes, just over two decades since our joint work on the new
Schürer began.
It could hardly be disputed that if we could recover exactly what was said
and done, around the time of Passover in an indeterminate year,1 to bring
about the crucifixion, the results would be of almost limitless importance.
But no such claim will be made here. Nor will the discussion take detailed account of the endless ‘‘bibliography of the subject.’’ 2 Instead, the emphasis will
* First published in P. R. Davies and R. T. White, eds., A Tribute to Geza Vermes: Essays on
Jewish and Christian Literature and History ( Journal for the Study of the Old Testament, supp.
ser. , ), –.
. I do not wish to enter into this question, but draw attention to the powerful converging arguments advanced for Passover of ..  by N. Kokkinos, ‘‘Crucifixion in .. : The
Keystone for Dating the Birth of Jesus,’’ in J. Vardaman and E. M. Yamauchi, eds., Chronos,
Kairos, Christos: Nativity and Chronological Studies Presented to Jack Finegan (), .
. I should note here the use I have made over the years of A. Wikenhauser, Einleitung
in das Neue Testament 5 (), and D. Guthrie, New Testament Introduction (), and more
recently L. T. Johnson, The Writings of the New Testament (). On the specific question
of the trial narratives, see especially E. Bickerman, ‘‘Utilitas crucis,’’ RHR  (): 
Studies in Jewish and Christian History III (), ; A. N. Sherwin-White, Roman Society
and Roman Law in the New Testament (); P. Winter, On the Trial of Jesus 2 (); O. Betz,
‘‘Probleme des Prozesses Jesus,’’ ANRW II, . (), .


Rome and the East
be, first, on examining the general characteristics of the Gospels, viewed as
biographical narratives (which is what they are, however ‘‘kerygmatic’’ their
intentions). This discussion will suggest some reasons why, if any one of the
Gospels can bring us closer to the historical context and overall pattern of
Jesus’ activities than the others, it is John rather than any of the Synoptics;3
while, of the Synoptics, it is Luke who has the weakest grasp on the realities
of Palestine under Roman domination. It is essential to stress that those realities provide the only touchstone for what may be veridical in any of the trial
narratives, as in the Gospel narratives as a whole. These realities are genuinely accessible, to a significant degree, because—and only because—of the
works of Josephus. In the case of Josephus we know who he was, what his
place was in the Jewish history of his time, what he wrote, when, where, and
to a large degree why. Not one of those questions can be answered with any
confidence for any one of the Evangelists. None the less, it is highly relevant
to note that Josephus’ Jewish War, Antiquities, and Life were themselves written in Rome in the s, s and s; a work can truly spring from the Judaea
of before ..  without having been written either there or then.
The evidence of Josephus enables us to say not which of the Gospel accounts is ‘‘true,’’ but, firstly, what is significant about the differences between
them; and, secondly, which of the things narrated by them could have been
true, and conversely which could not. To take one example: the two birth
narratives, of Matthew and Luke, are wholly different, and mutually incompatible; but Matthew’s account fits with historical reality and could be true in
its broad outlines, while Luke’s does not, and cannot be true. This distinction does not lose its significance even if we conclude, as I believe we must,
that in fact neither is true.
If we then turn to the trial narratives themselves, we may be able to find
reasons why some are likely to be false, because they do not fit with what we
know from more secure evidence. And we may also be able to show that one
is plausible, that it does ‘‘fit.’’ But that is not the same thing as proving it to be
true. For it lies in the nature of arguments from coherence that we can never
confidently distinguish between an essentially veridical narrative, based on
first-hand reports, and a convincing reconstruction—or fiction—whose author respected historical realities. We cannot know ‘‘what happened’’; but we
can certainly gain a clearer idea of the significance of the differences between
the several accounts we are given.
. My use of, and emphasis on, John clearly owes much to C. H. Dodd, Historical Tradition in the Fourth Gospel (), and to J. A. T. Robinson, The Priority of John (). See now
M. Hengel, The Johannine Question ().
Reflections on the Trials of Jesus

That we are given quite different accounts is, of course, well known. For a
start, in the Synoptics the Last Supper is a Passover meal at which the paschal
lamb is eaten, and in John it is merely a meal on the evening before Passover. We may not be able to prove which, if either, of these versions is true,
though some reasons will be advanced below for preferring John’s version.
But what is logically beyond dispute is that they cannot both be historically
true; and therefore that at least one of them is false. I make no apology for
placing so much weight on the question of literal, non-metaphorical, nontheological, mundane truth or falsehood; for that after all is what historians
are for.
The Gospels as Biography and History
Before offering a view on the fundamental question of historical truth, it is
essential, however briefly, to look at the Gospels overall when considered as
biographical narratives. I assume in the following discussion that the conventional view that Mark’s Gospel is the earliest of the Synoptics is correct;
but also that Matthew represents a development of Mark, and that Luke has
probably used both; and that John is independent. The trial narratives themselves, which represent so prominent a part of the structure of each of the
Gospels, lend strong support to these views.
I take up no position on the sources of the Gospels, or on the question
of the absolute date when any one of them was written, or even on whether
they were written before or after the siege of Jerusalem and the destruction
of the Temple in .. . All that seems to me to be certain of all four is
that they could not have come to be as they are without their deriving in
some sense, direct or indirect, from an environment in which the geography
and social structure of pre-..  Palestine was familiar; and, more important, an environment in which the concerns of pre- Jewish society were
still significant, whether we think of the high priests and ‘‘the Sanhedrin,’’
of Pharisees and Sadducees, of the relations between Galilee, Samaria, and
Judaea, or of the centrality of the Temple and of pilgrimage to it, and of the
major festivals celebrated there, Passover above all. In a profound sense, the
world of the Gospels is that of Josephus. But there remains one major puzzle,
to which too little attention has been directed: in the Synoptic Gospels two
groups called grammateis (scribes) and presbyteroi (elders) play a major role. But
these terms, in the plural, as designations of apparently definable groups, are
unknown to Josephus’ accounts of the period, in the War and the Antiquities; it may therefore be significant that they are also unknown to John. With
that exception, and allowing for very considerable variations between them,

Rome and the East
the Gospels all ‘‘belong’’ in pre-..  Palestine, and must in some sense derive from it. Within that wider framework it must be firmly asserted that
the Gospels are biographical narratives; Matthew and Luke follow the life of
Jesus from birth to death; Mark and John do so from his recognition by John
the Baptist until death. The two pillars on which the structure of all four
narratives rest are therefore, firstly, John the Baptist and his proclamation of
Jesus and, secondly, the Passion narratives.
Only Matthew and Luke take the story back to the birth of Jesus, and
do so in wholly different and incompatible ways. But we cannot understand
the significance of this comparison unless we hold fast to the historical framework of the later years of Herodian rule and the early stages of Roman rule
in Judaea, as provided by Josephus.4 If we use this framework, we find that
Matthew presents an entirely feasible succession of events. Jesus was born
in Bethlehem of Judaea in the days of Herod the King (:), therefore not
later than spring of  .. In fear of Herod, Joseph took the family to Egypt
(:), which since  .. had been a Roman province. When Herod died
( ..), an angel prompted Joseph to return to the land of Israel; but hearing that Herod’s son Archelaus was ruling in Judaea, as he did from  .. to
.. , Joseph was afraid. So he went instead to Galilee, and settled in Nazareth (:). The implication is that Archelaus was not ruling there, which is
correct. It is not, however, explicitly stated at this point that the ruler there
was Archelaus’ brother, Herodes Antipas, who was in fact in power there
from  .. to .. . Indeed it is not until :– that ‘‘Herodes the tetrarch’’
(his correct title) makes his first and only entrance, with a reference back to
his execution of John the Baptist. So the historical framework is only partially reflected; all the same the underlying presumption that there was more
to fear in Judaea under Archelaus than in Galilee under Herodes Antipas is
borne out by Josephus’ accounts of the two reigns.
The purpose of the story is to explain how Jesus, later to emerge from obscurity as a man from Nazareth, both belonged to the line of David (hence
the genealogy in :–) and had in fact duly been born in Bethlehem (:
–).
Luke’s birth narrative has the same purpose, but sets about fulfilling it
quite differently. Even his genealogy, which he does not introduce until
chapter , disagrees with Matthew’s, beginning in the generation before
Joseph; but it too includes King David (:–). More important, having
begun by locating the story ‘‘in the days of Herod the king of Judaea’’ (:),
. For the historical framework and datings, and above all the crucial question of the
census of .. , see Schürer, Vermes, and Millar, History I, –.
Reflections on the Trials of Jesus

he continues with the episode of Zechariah and Elizabeth, coming only in
: to Mary and her fiancé Joseph, ‘‘from the house of David,’’ but settled
in Nazareth in Galilee. The birth of Jesus in Bethlehem is brought about by
the proclamation of the census, requiring all to go to be registered, each to
his own city (:). So Joseph and Mary go to Bethlehem ‘‘since he belonged
to the house and kindred of David.’’
Unfortunately the story is a historically impossible construct, which
makes use of the long-remembered and traumatic moment when in .. ,
ten years after Herod’s death, and following the deposition of Archelaus,
Judaea became Roman provincial territory, and the Roman census, a complete novelty, was imposed. A resistance movement flared up, and was repressed with difficulty. Luke is quite unaware of this precise context. But he
has also forgotten something much more significant. Neither in ..  nor
at any other time in the life-time of Jesus was Galilee under Roman rule, or
subject to the census. Furthermore, as we know from a much-quoted papyrus of .. , the Roman census in fact required people to return, not to
their ancestral home, but to their normal place of work and residence, which
in the case of Joseph would have been Nazareth.5
We need not pursue the argument further. Both birth narratives are constructs, one historically plausible, the other wholly impossible, and both are
designed to reach back to the infancy of Jesus, and to assert his connection
with the house of David (as it happens, almost the only characteristic of the
earthly Jesus alluded to by Paul, Rom. :) and his birth in Bethlehem. For
if it could be known at all from where the Christos would come (for some
doubts on this see Jn. :), then it ought surely to have been Bethlehem;
the expectation is underlined most clearly of all in John (:–): ‘‘And they
said, ‘Surely the Christos does not come from Galilee? Has not Scripture said
that it is from the seed of David, and from Bethlehem, the village where
David was, that the Christos comes?’ ’’ John does not claim that in this respect prophecy had been fulfilled; and it is he alone of the Evangelists who
confronts this failed expectation.
This is not the place to attempt to examine in detail the different accounts
in the Gospels of the various episodes of Jesus’ ministry between his recognition by John and his journey to Jerusalem, arrest, ‘‘trial,’’ and crucifixion.
Such a detailed discussion would serve no purpose, for, as mentioned above,
all four Gospels show every sign of deriving, directly or indirectly, from the
real historical environment of Jesus’ preaching, in Galilee, Peraea, in the territory of Caesarea Philippi, and of Tyre and Sidon, and en route between
. P. Lond. , lines –; Sel. Pap. II, no. .

Rome and the East
Galilee and Jerusalem. But it is essential to state firmly that we cannot amalgamate the four accounts to construct a ‘‘life of Jesus.’’ We could attempt to
do so with the three Synoptics, but not with John, because the structure of
his narrative is fundamentally different. For the Synoptics, there is only one
journey to Jerusalem, that for the final Passover, the occasion of the crucifixion. Their narratives thus lead from Galilee and its environs to a single
climax, namely the one pilgrimage to Jerusalem for Passover. As such, this is
entirely convincing. Josephus’ two narratives of the period give ample evidence that Passover was indeed the main national pilgrim festival, when vast
crowds assembled, from Galilee not least, and when disturbances could be
anticipated.6
This concentration on a single climactic visit has its effect also on the details of the Synoptic narratives. So, for instance, in Jesus’ life-time the Roman
census was imposed, and Roman taxation was payable, in Judaea but not in
Galilee, a fact which, as we have seen, Luke’s birth narrative overlooks. The
question of payment remained a burning issue. So all three Synoptics represent the trick question about whether to pay ‘‘the census’’ as having been
posed in Jerusalem, necessarily in the period before the last Passover (Mk. :
–; Mt. :–; Lk. :–). It is also in this context that they must
place the cleansing of the Temple (Mk. :–; Mt. :–; Lk. :–
). But in John this episode belongs in a quite different context. For just
as his trial narrative is structured round Pilate’s movements between Jesus,
inside the praetorium (the residence of the Roman praefectus), and his Jewish
accusers outside, as we shall see, so his narrative of Jesus’ preaching is structured round a whole series of Jewish festivals, proceeding in what looks like
an appropriate sequence through at least something more than one year, and
each necessitating an ascent from Galilee to Jerusalem. The sequence begins
with a first Passover, almost the earliest episode in Jesus’ activity as a preacher,
being preceded only by the marriage at Cana (:–), an item unique to
John. (Cana was a real village in Galilee, where Josephus once stayed on campaign in ..  [Vita ]; unfortunately he does not report having heard
there any interesting local tales.) It is thus very early in the narrative that
John represents Jesus as going up to Jerusalem for ‘‘the Pascha of the Ioudaioi,’’
cleansing the Temple, meeting ‘‘a man of the Pharisees, Nicodemus by name,
an archon [magistrate] of the Ioudaioi,’’ who is to reappear later after the crucifixion (:), and then going out into the countryside of Judaea (:–:
). On his way back he has to pass, as Galilean pilgrims often did,7 through
. See, e.g., BJ , –; , –; , –; , –; Ant. , –; , –.
. See, e.g., BJ , –; Ant. , –.
Reflections on the Trials of Jesus

the territory of Samaria. The picture then offered of Samaritan beliefs and
attachment to their sacred mountain (Mount Gerizim) is the most detailed
in any of the Gospels (:–) and is vividly matched by Josephus’ description (Ant. , –) of an episode which belongs very soon after the time of
Jesus’ preaching: a local man persuaded a large group to ascend Mount Gerizim in the hope of finding there sacred vessels buried by Moses; but the
movement was suppressed by a force sent by Pilate. The episode is followed
by the dismissal of Pilate, apparently in the winter of .. /.
In John’s narrative Jesus now returns to Galilee, but his work is interrupted
by a ‘‘festival of the Ioudaioi’’ (:), perhaps Pentecost (though the words could
refer to the Passover of the next year), and Jesus goes up again to Jerusalem,
where he heals a lame man lying at the pool called Bethesda and is blamed
for doing so on a Sabbath (:–). After his reply, and with no transition,
he is found going away across ‘‘the sea of Galilee to Tiberias’’ (:, with :
the only reference to this new city in the Gospels). He then ascends a mountain; but here the chronological sequence may be in some way distorted, for
John says (:) that ‘‘the Pascha, the festival of the Jews’’ was near. There is
no further reference to Passover in the long section which follows (:–).
This must be either a displacement, or a scribal or authorial error, or we have
shifted forward a whole year (or even two years altogether, if the two allusions to a ‘‘festival of the Jews’’ [: and :] should be taken as referring to
two successive Passovers). But to solve the problem in that way would be to
indulge in an inappropriate literalism. It is perhaps more likely that there is
some mistake here, and that John is intending to portray Jesus’ preaching and
his movements from Galilee up to Jerusalem and back within the framework
of a cycle of festivals covering just over one year. If that is so, then it is appropriate that we come in :– to Sukkot/Tabernacles. Jesus is preaching
in Galilee, for he fears to go to Judaea. But his followers urge him to go none
the less, for ‘‘the festival of the Jews, skenopēgia [Sukkot]’’ is approaching. Jesus
then does go, first clandestinely, then teaching openly in the Temple during
the festival. No features of this festival are explicitly reflected in this section
(:–), except for an allusion to the last day, as being the climax (:); but
the atmosphere of a popular festival centred on the Temple is felt throughout. It is worth recalling that, in a year which is probably .. , almost
contemporary with Jesus’ preaching, the Jews of Berenice in Cyrenaica held
an assembly at which they voted honours to a Roman official ‘‘at the time of
the public assembly [syllogos] of the skenopēgia’’ (IGR I, ).
As was mentioned above, we cannot regard the fact that the Synoptics
represent Jesus as going up to Jerusalem only for one Passover as itself a strong
argument for preferring John; for it is indisputable that Passover was indeed

Rome and the East
the major national pilgrim festival. None the less John’s representation of
Tabernacles as being a strong reason for going up to Jerusalem clearly accords with our other evidence. Josephus emphasises the special importance
of the festival (Ant. , –) and the requirement to go up to Jerusalem
to celebrate it for eight days (Ant. , –); but, apart from the rules for
the festival given by Philo,8 perhaps the most striking of all testimony to
its significance in this period is Josephus’ casual reference to the fact that in
..  Cestius Gallus was able to capture Lydda with ease, because the entire
population had gone up to Jerusalem for the feast of Tabernacles (BJ , ).
It is also in the context of this celebration of Tabernacles that John represents differences breaking out over the inappropriateness of a Christos coming
from Galilee, and not from Bethlehem; and it is also here that the arrest
and examination of Jesus is first concretely foreshadowed, for their attendants report about him to the archiereis (high priests) and Pharisaioi, who reproach them for not bringing Jesus with them. We may recall a profoundly
relevant episode which, as Josephus records (BJ , –), took place at
Tabernacles in .. , when a peasant, also named Jesus, began to prophesy in public against Jerusalem and the Temple, was arrested by the archontes
(magistrates), brought before the then Roman governor, Albinus, flogged,
cross-questioned, and finally released as a lunatic.
In John’s narrative there follows a long section representing Jesus’ preaching and healing, with no explicit indication of his either staying in Jerusalem
or leaving it ([:–]; :–:). But then there comes a festival described
as ta enkainia (festival of restoration), in Jerusalem. It can only be Hanukkah,
for the meaning, ‘‘renewal,’’ is the same, and we are in winter (:); Jesus
spends his time at the Temple, in the ‘‘Stoa of Solomon.’’ We know from
Josephus that this festival was celebrated for eight days but not how it was
celebrated, except (in general terms) with hymns, sacrifices, and popular rejoicing (Ant. , –). Jesus again runs into danger, this time the threat
of stoning by the mob, and escapes. He then leaves Jerusalem to cross the
Jordan to the spot where John had first baptized him, and is called back to
raise Lazarus from his tomb at Bethany near Jerusalem (:–:). It is
this miracle which provokes an initial synhedrion (meeting) of the high priests
and Pharisees, at which they resolve to put Jesus to death for fear of the Roman reaction if all the people were to follow him. Caiaphas makes his first
appearance, described as ‘‘archiereus [high priest] of that year,’’ and utters the
proposition that one man should die for the sake of the people. This thought,
. Philo, De Spec. Leg. , ; , –.
Reflections on the Trials of Jesus

John says, was in reality an inspired prophecy; for Jesus would die not only
on behalf of the people, but so that he might gather together the scattered
children of God (:–). The last Passover is approaching, and large numbers are going up to Jerusalem in advance of it in order to purify themselves
(:). Meanwhile Jesus goes off to the desert, then to Bethany again, and
then makes his formal entry to the city (:–:).
What is beyond question is, first, that John presents to the reader an incomparably more detailed and circumstantial picture of Jewish life in Palestine, punctuated by the annual rhythm of the festivals, than do the Synoptics.
He also represents the earthly activity of Jesus as reaching a series of preliminary climaxes in Jerusalem, in close association with these festivals. As
indicated above, there is no way, logically speaking, in which we can distinguish between a ‘‘true’’ narrative and one which is plausible and evidently
related to a historical and social framework known from other evidence. On
the one hand, either we must abandon altogether even the attempt to decide
questions of literal truth; or we must conclude that nothing in any Gospel
has any claim to truth; or we must choose. Either Jesus went only once to
Jerusalem, for the fatal Passover, or he went several times, for a succession of
festivals. Either we have no evidence at all which offers us any access to the
earthly life of Jesus, or we must choose between John and the Synoptics. The
only criterion of truth in the Gospels which a historian can offer is conformity with the world as portrayed by Josephus, and what we have in John may
be no more than a convincing fiction. But, as we must choose, I suggest that
the narrative of Jesus’ ministry which brings us closest to the real world of
first-century Palestine is that of John.
This suggestion can be no more than that. It is not only that we cannot
prove, or disprove, the literal truth of any statements in the Gospels, or demonstrate the validity of one narrative structure as against another. Nor can
we ever escape from our inability to distinguish a true story from a convincing literary construction. But since we can hardly fail to wish to know more
about these events, whether we see them as embodying a divine revelation,
or merely the most important single turning point in world history, we both
may and should form hypotheses about where we should begin in thinking
about the life of Jesus. All of the available accounts give a very large place to
the story of how he came to be crucified. I will therefore suggest that if we
wish to consider these narratives, we should be open to the hypothesis that
the one to which we should give preference is that of John.

Rome and the East
The Trial Narratives
As already suggested, the trial narratives occupy a central place in the structure of all four Gospels. Both in scale and in coherence they have to be taken
as representing a significant aspect of what the Gospels, conceived of as narratives, are. It is also quite possible, though it cannot be proved, that they
represent the earliest narrative sections to come into existence; on this hypothesis the Gospels, as biographical narratives, will have grown backwards.
It is noticeable, as we have seen, that only two of them stretch back to Jesus’
birth, both in unconvincing ways, though Luke much more unconvincingly
than Matthew.
As is well known, the trial narratives also present profound and irreconcilable differences, Mark/Matthew from Luke and, much more profoundly,
all three Synoptics from John. The differences centre both on the timing of
the Last Supper and the crucifixion, and, in ways which need more emphasis
than they have received, on the significance of Passover as a factor which determines how the events unfold. In the Synoptics the Last Supper is a paschal
meal eaten on the first night of Passover, the examinations of Jesus take place
during that night and in the following morning, and the crucifixion follows
on the first day of the festival. In John all this happens one day earlier, and
the beginning of Passover, on the evening of the day of the crucifixion, is
still expected.
Not all the features of the celebration of Passover as it was in the first century .., while the Temple still stood, need to be considered here; and many
aspects in any case remain somewhat obscure.9 But certain points are crucial.
First, Josephus makes clear that the people would begin to assemble some
six days before the festival, on the th of Nisan (BJ , ); we have already
seen this reflected in John’s narrative when ‘‘many went up to Jerusalem from
the country before the Pascha, so that they might purify themselves’’ (:).
Just after this a precise date is given: six days before the Pascha Jesus goes to
Bethany (:).
Passover proper began on the evening of th Nisan, though it seems clear
that to accommodate the enormous number of private sacrifices now offered
by the people in groups, the long sequence of sacrifices had in fact moved
back into the afternoon of that day, long before sunset.10 Indeed the festi. I rely on the very interesting, if not always entirely clear or conclusive, discussion by
J. B. Segal, The Hebrew Passover from the Earliest Times to ..  ().
. For the conduct of the sacrifices in daylight, during the afternoon of th Nisan,
see Segal (n. ), , using Jub. . For the process of sacrificing in groups, amounting to
Reflections on the Trials of Jesus

val cast its shadow even further back, for all of th Nisan counted as a day
of preparation;11 as we will see, John twice described this day as paraskeuē
(preparation, : and ). He may also single out the first day of the festival,
when the pascha was eaten, from the (originally separate) days of unleavened
bread which followed. That depends on how we read a crucial phrase in :
(see below). But in any case the later phase of Passover need not concern us.
What is, however, crucial is the question of purification, stressed (see below)
both by Ezra and by Philo, and apparently affecting all participants, and not
just priests. In the same passage in which Josephus gives the total of participants at Passover of (probably) .. , he too stresses the exclusion of anyone
suffering from any form of impurity.12
These broad principles, which clearly ignore many finer points, may provide a sufficient framework for understanding the highly significant differences between the various Gospel narratives.
Mark
Mark’s account begins with a Last Supper which takes place on the first night
of Passover and has as its purpose the eating of the pascha (:–), and continues with the arrival of Judas at Gethsemane, accompanied by an armed
mob ‘‘from the high priests [archiereis] and the scribes [ grammateis] and the
elders [ presbyteroi]’’ (:). Jesus is then led to the high priest (archiereus), and
all the high priests, elders, and scribes assemble. The scene is the house of
the high priest, where Peter is warming himself. Inside, the high priest and
‘‘all the council [synhedrion]’’ hear testimonies against Jesus in order to kill
him (:). Two questions are specifically addressed to Jesus, about his proclamation that he would destroy the Temple, and about his claim to be the
Christos, which Jesus admits, emphasising his claim with a quotation from
Daniel (:).
vast numbers of individuals in all, see Josephus, BJ , –. On Josephus’ account, at Passover of (apparently) .. , , victims were sacrificed on behalf of groups of ten to
twenty each, these sacrifices being carried out between the ninth and the eleventh hour,
hence towards evening, but before sunset.
. For this point, cf. Segal (n. ), , which does not however cite any very clear evidence. The term paraskeuē is attested as meaning ‘‘the day before Shabbat,’’ e.g., Josephus,
Ant. , , but so far as I know not elsewhere unambiguously in relation to a festival. But
(see below) Jn. : uses ‘‘paraskeuē of the Pascha.’’
. BJ , –: ‘‘all pure and holy. For those afflicted with leprosy or gonorrhoea, or
menstruating women, or persons otherwise defiled were not permitted to partake of this
sacrifice.’’ See S. Safrai, ‘‘The Temple,’’ in Jewish People in the First Century II, ff., on –.

Rome and the East
This scene takes place at night, and at dawn Peter makes his denial, and
a cock crows (:–). Immediately in the morning, having taken council, the high priest with the elders and scribes and the whole council bring
Jesus bound before Pilate, who asks him a different question, ‘‘Are you the
king of the Jews?’’ The high priests make further accusations, receiving no
reply (:–). There follows an episode involving the custom of releasing
a prisoner on the occasion of Pesach, the crowd’s demand for the release of
Barabbas, ‘‘imprisoned with his insurgents [stasiastai], who had committed
murder in the uprising [stasis],’’ Pilate’s dialogue with the crowd, its demands
for the crucifixion of Jesus, the release of Barabbas, and the delivery of Jesus
for crucifixion. Jesus, now apparently outside, is taken within the house or
praitōrion ( praetorium), abused by the soldiers, and led off (:–). Simon
the Cyrenaican is commandeered en route, the procession reaches Golgotha,
and the crucifixion takes place at the third hour, the cross being inscribed
‘‘The king of the Jews,’’ in what language is not stated (:–).
The whole account, from the arrest to the inscription on the cross, occupies fifty-six verses, or a little over one chapter. It involves two examinations
of Jesus, one at night in the house of the high priest and one in the early
morning in the residence, or praitōrion, of Pilate; but it represents no formal
trial or verdict. The phrase which Mark uses of deliberations in the morning, symboulion poiēsantes (:), might indeed be read as meaning ‘‘having
held a council meeting’’; but where he uses it elsewhere it means no more
than ‘‘took counsel’’ or even ‘‘conspired’’ against Jesus (:). Pilate’s order for
crucifixion is prompted by the demands of the crowd.
Matthew
Matthew’s account has an almost exactly similar structure, beginning with
a Last Supper for the eating of the pascha (:–). The mob which arrests
Jesus, however, comes from the high priests and elders of the people (laos).
The high priest to whose house Jesus is brought is identified as Caiaphas, and
this time the high priests are omitted from the list of those who assemble
there (who are described as ‘‘the scribes and the elders,’’ :). None the less,
those reported as seeking false testimony against Jesus are then described as
‘‘the high priests and the whole council’’ (:). The reported dialogue is
closely similar, and it is followed again, at dawn, by the taking of counsel by
the high priests and the elders of the people, who bring Jesus bound before
Pilate (:–). After a complete inserted episode relating to Judas and the
thirty pieces of silver (:–), Jesus appears before Pilate and is again asked,
‘‘Are you the king of the Jews?’’ The episode of Barabbas follows, with the
Reflections on the Trials of Jesus

extra detail of Pilate’s being seated on his tribunal (:), with the anecdote of his wife’s dream, and the detail of his washing his hands (derived
from Deut. :–). The rest follows as in Mark, except that the reported
inscription on the cross is longer: ‘‘This is Jesus, the king of the Jews’’ (:).
The narration is thus a little fuller, with extra details, and occupies sixtyfour verses, or a matter of a chapter and a half. The structure is identical: two
examinations, neither in the form of a trial concluded by a verdict, and an
order for crucifixion prompted by the demands of the crowd.
Luke
Luke’s account similarly follows a Last Supper which involves the eating
of the paschal lamb (:–), and begins with the arrival on the Mount
of Olives of Judas with a ‘‘crowd’’ (ochlos), not otherwise identified (:).
When Jesus addresses them, however, they turn out to be (or to include) high
priests, captains (stratēgoi) of the temple, and elders. They take Jesus to the
house (oikos) of the high priest, who is not named. Peter’s denial follows (:
–), but the structure of the narrative then becomes crucially different.
For when dawn breaks a formal council of the people (to presbyterion tou laou)
is convened: ‘‘high priests and scribes, and they took him to the synhedrion’’
(:). The shift is crucial in two different ways. Firstly, Luke transfers to
here Jesus’ reply to the question as to whether he is the Christos, and his answer
quoting Daniel. Secondly, this passage is the only one in the four Gospels
which seems to represent a formal meeting of the body normally known in
modern literature as ‘‘the Sanhedrin.’’ This concept has its problems, as we
will see below; and the term presbyterion is used of ‘‘the Sanhedrin’’ only by
Luke himself (otherwise in Acts :). However, Luke clearly intends to differentiate between an examination at night in the house of the high priest
and some sort of formal meeting of a council in the morning. None the less,
even here, no concluding verdict of the meeting is represented.
Luke continues by specifying, as neither Mark nor Matthew does, exactly
what accusations were put forward when Jesus was brought before Pilate:
‘‘We have found this man disturbing our people, preventing them from giving tribute to Caesar and calling himself Christos Basileus’’ (:). However,
Pilate asks the same question, ‘‘Are you the king of the Jews?’’ Luke then gives
a unique twist to the story by having the high priest (and the crowd?) say
that Jesus has been upsetting the people, teaching throughout Judaea, beginning from Galilee. This prompts Pilate to ask if Jesus is a Galilean, and, on
discovering that he is, to send him for examination to Herodes (i.e., Herodes
Antipas, the tetrarch of Galilee and Peraea), who happens to be in Jerusalem.

Rome and the East
Herodes examines him, and sends him back to Pilate (:–). There is no
inherent improbability in Herodes’s presence; Agrippa II, when ruling part
of Galilee, and other areas, but not Judaea, was to maintain a palace in Jerusalem, and came there frequently ( Josephus, Ant. , –). None the less,
Acts provides a very clear indication of how this episode came to be added
by Luke, and by him alone: for in Acts :– the early Christian community is found quoting Psalm :–: ‘‘The kings of the earth set themselves,
and the magistrates were gathered against the Lord [Kyrios] and against his
anointed [Christos],’’ and applying this to the double examination of Jesus
before Herodes and Pilate. As is notorious, the fact that an episode in the
Gospels is explained or justified in terms of a biblical quotation does not necessarily prove that the episode concerned is invented. But the presence of
this episode and its re-emphasis in Acts serves at any rate to underline the
freedom of Luke’s use of whatever material he had before him.
In the Gospel Luke returns to Pilate’s dialogue with the high priests and
magistrates of the people, whom he summons for a second meeting, and duly
refers to Herodes’s inability to find Jesus guilty. With that variation, the exchange leads on to another narrative of the dialogue involving the release
of Barabbas, brought in without any explanation of the custom or its relation to Passover, and ending with the delivery of Jesus for crucifixion. The
taunting of Jesus by the soldiers is omitted, but Luke chooses to explain that
Simon the Cyrenaican was ‘‘coming from the field’’ (:), a bit of narrative
colour which however sits unconvincingly with the idea that this is the first
morning of Passover. The action moves to ‘‘the place called Kranion’’; the
inscription on the cross is given almost as in Mark, ‘‘the king of the Jews,’’
again with no indication of the language used (:).
The scene before Herodes, unique to Luke, remains a puzzle, and its inauthenticity certainly cannot be demonstrated. But the crucial variation is the
representation of a formal council meeting in the morning, reinforced by a
very concrete later reference to Joseph of Arimathea: ‘‘he happened to be a
councilor [bouleutēs], a good man and just (he did not agree with their counsel or deed)’’ (:). Mark had indeed referred to him as a counsellor (:
)—though Matthew only as a rich man (:)—but had necessarily not
deployed any allusion to his non-participation in the relevant deliberations.
All the endlessly debated questions as to whether ‘‘the Sanhedrin’’ had the
formal right to pass a sentence of death and, if so, whether it was compelled
to have that sentence carried out by the Roman governor can thus be relevant to the Gospels only in relation to Luke’s Gospel; for it is only here that
something which is clearly a meeting of ‘‘the Sanhedrin’’ is represented as
taking place. Luke’s narrative might even gain some support from the pro-
Reflections on the Trials of Jesus

vision in Mishnaic tractate Sanhedrin, that that body could only meet as a
court in the hours of daylight (:). Appeal to the Mishnah will hardly help,
however. For the same passage also lays down that a capital trial could not
be held on the day preceding a Sabbath or a festival day. Even disregarding
the insoluble question of whether the Mishnah preserves any veridical conception of how justice was exercised before the destruction of the Temple,
the notion that an arrest, an examination in the house of a high priest, perhaps also a formal meeting of ‘‘the Sanhedrin,’’ the production of a prisoner
before the Roman praefectus, popular demands for execution, and the crucifixion itself could all have taken place on the night of Passover and on the
following morning must give rise to serious questions. It is time to turn to
John’s account.
John
As is well known, the overall structure of John’s narrative differs fundamentally from that which is common to the Synoptics, and those details which
do reappear in John mainly do so in a quite different narrative context.
It was mentioned before that the preceding narrative of the Last Supper
explicitly locates the event before Passover (:), and the lengthy account of
it is consistent in betraying no trace of its having been a paschal meal (:–
:). This location in time is to be fundamental to the logic of the story,
as we shall see. Further differences also appear immediately. For the armed
band which Judas leads out to arrest Jesus in the garden beyond the Kedron
valley is composed of the speira and attendants of (or sent by) the high priests
and the Pharisees (:). Speira is the normal Greek translation of cohors, and
the impression that this is intended to be understood as a Roman detachment
is confirmed by the reference a little later to its commander as the chiliarchos,
the normal Greek for tribunus (cf. the chiliarchos of the speira of Acts :, who
turns out to be the Roman officer Claudius Lysias). No explanation is given
by John of this Roman involvement, and indeed no comment is made on it
at all. The combined Roman-Jewish group brings Jesus first to (the house of )
Annas, a person not mentioned in any other Gospel in this context, but carefully identified here: ‘‘he was the father-in-law of Caiaphas, who was high
priest of that year’’ (:). John further identifies Caiaphas by saying that it
was he who advised the Jews that it was advantageous for one man to die
on behalf of the people (:), explicitly referring back to :–, where
the advice had been recorded in the context of a council (synedrion) of high
priests and Pharisees, and Caiaphas had already been identified as ‘‘the high
priest of that year.’’ If John meant to imply that the high priesthood changed

Rome and the East
every year, he was of course wrong. But he was correct in that Caiaphas
was indeed high priest from about ..  to . His father-in-law Annas or
Ananus (the relationship is attested only here) had, however, also earned the
designation high priest by holding the high priesthood from ..  to .13
The term is thus also applied (by implication) to him when Peter and another
disciple follow Jesus into ‘‘the house [aulē ] of the archiereus’’ (:), where
Peter makes his first denial. A first examination of Jesus is then conducted
by the high priest, evidently Annas, leading to no clear answer, after which
Annas sends him bound to appear before Caiaphas the high priest (:–
). It is here that Peter is again described as warming himself and making
his second denial (:–).
No examination in the house of Caiaphas is represented. There is thus a
clear contradiction of all three Synoptic narratives, which represent Jesus as
being brought to only one high-priestly house, where the examination is
conducted; and specifically of Matthew, who names the high priest as Caiaphas. No taking of counsel in the morning is recorded either, and instead
‘‘they took Jesus from [the house of ] Caiaphas to the praitōrion.’’ The time is
indicated clearly, and with it both the crucial element in the whole account
and a source of considerable difficulty: ‘‘it was morning, and they did not
enter into the praitōrion, so that they might not be defiled, but (be able to) eat
the pascha’’ (:). The context is thus made quite clear, even if the nature of
the defilement which would ensue on entering the praetorium of the Roman
praefectus is not immediately obvious. It is, however, clear, though not explicitly stated, that the group which would otherwise have entered the praetorium, coming as representatives directly from the house of the high priest,
should be presumed by the reader to have included at least some kohanim
(priests); this indeed becomes explicit later (:). Whether an even more
profound significance is to be attached to this precise moment in time depends on how we understand the exchange which then follows between
‘‘them’’ (not identified in any way at this point) and Pilate. He comes out of
the praetorium to meet them and asks what charge they are bringing. They
reply that, if Jesus were not a wrongdoer, they would not have handed him
over. Pilate then replies, ‘‘Take him yourselves and judge him according to
your own law.’’ They (here defined simply as the Iudaioi) say, ‘‘It is not permitted to us to kill anyone.’’ John’s authorial comment is that the purpose of
this was to ensure the fulfilment of Jesus’ own prophecy as to what sort of
death he would suffer (:). The comment repeats exactly what John had
. For the family, see now D. Barag and D. Flusser, ‘‘The Ossuary of Yehoḥanan Granddaughter of the High Priest Theophilus,’’ IEJ  (): .
Reflections on the Trials of Jesus

said at :–, referring to Jesus’ prediction that he would be ‘‘raised up’’
(i.e., on a cross).
Leaving aside for a moment the question of how we should understand the
exchange between Pilate and the Jews, it should be stressed that the structure
of the narrative which follows depends entirely on the physical separation
of Jesus, under arrest in the praetorium, and the Jewish group outside. In this
narrative Pilate now goes back inside and questions Jesus about his alleged
claim to be a king (basileus) (:–), and then comes out again to offer the
release of Barabbas. John relates the custom to Passover, but puts the explanation of it in the mouth of Pilate addressing the Jews. Barabbas is identified
by John simply as a lēstēs, a brigand (:–).
The episode of the mocking of Jesus, still in the praetorium, follows (:–
), after which Pilate comes out again and tells those outside that he can find
no cause for accusation in Jesus. Jesus then himself comes out (or is brought
out, as must be understood), wearing the mock royal crown and robe. When
Pilate displays him to those waiting, they, described as the high priests and
the attendants, shout ‘‘crucify! crucify!’’ Pilate then ‘‘seeks to release him,’’ by
which it seems to be implied, though it is not stated, that he has gone outside to speak to the Ioudaioi again. For they then shout, ‘‘If you release him
you are not a friend of Caesar, for anyone who makes himself a king is an
opponent of Caesar.’’
At this point John’s narrative again goes in for deliberate and emphatic detail as regards place and time. This has to be understood as a significant feature
of it as narrative; no presumption can follow as to whether these details are or
are not historically valid. Pilate’s response to these shouts is to lead Jesus out
again and take his seat on his tribunal (bēma) ‘‘in the place called Lithostrōtos,
but Hebraïsti ‘Gabbatha’ ’’ (here as elsewhere it is not clear whether Hebrew
or Aramaic is referred to by Hebraïsti). The time is given with equal precision: it was the day of preparation for the pascha, and the time was ‘‘about the
sixth hour.’’ The location is clearly understood to be a paved stone courtyard,
out of doors, and immediately outside the praetorium; it was in this courtyard that the regular tribunal from which the governor gave jurisdiction and
held audience was situated. Matthew also mentions the bēma/tribunal (:
), but does not give its location; and the consistent separation of inside,
with consequent impurity, and outside, in the open air, plays no part in the
Synoptic narratives. Nor of course does this indication of the date and time.
Once seated on his tribunal, Pilate responds to shouted demands for crucifixion by asking, ‘‘Shall I crucify your king?’’ and the high priests answer,
‘‘We have no king but Caesar.’’ John then appears to say that Pilate handed
Jesus over to them for crucifixion (:), and continues by saying that ‘‘they’’

Rome and the East
took him. But as the narrative unfolds it becomes clear that the execution,
and the division of Jesus’ clothing (fulfilling Ps. :), is being conducted
by Roman soldiers (:–). John’s detailed and concrete narrative style is
demonstrated to the end, though he makes no reference to Simon the Cyrenaican. Jesus is brought ‘‘to the place called (place of ) a skull, or, as is said
Hebraïsti, ‘Golgotha’ ’’; he thus reverses the equivalence stated by Mark (:
). The element which is central to all accounts of the crucifixion, the inscription on the cross, is given here in much more detailed form. First, a
longer version of the text itself is offered: ‘‘Jesus the Nazarene, the king of
the Jews’’ (:). John goes on to say that many Jews read the inscription
(titlon) since it was written Hebraïsti, Rōmaïsti, Hellēnisti (in Hebrew or Aramaic, in Latin, in Greek). The use of the Latin loan-word titlos (from titulus)
is unique to John among the four Gospels. So is the indication of the trilingual character of the inscription. John alone concludes his narrative with
an exchange between Pilate and ‘‘the high priests of the Ioudaioi,’’ who complain that the appellation ‘‘king’’ should have been set out as what Jesus had
claimed to have been, and not as the actual truth.
A step-by-step analysis of the structure of John’s narrative is essential if
we are to avoid the trap of attempting to amalgamate the separate accounts,
or of selecting convincing details from each, to make up a historical reconstruction of ‘‘what really happened.’’ John’s account is in no way compatible
with those of the Synoptics. It is not only that there are many different details, or even that the sequence of events unfolds quite differently. It is that
the overall setting of these events in relation to the Jewish calendar is different, and significantly different: the Last Supper does not have the character
of a paschal meal, and the nature of the exchanges between Pilate and the
Jewish leaders is entirely determined by their refusal to enter the praetorium
in order to avoid pollution, and thus not be prevented from eating the pascha
later that day.
It has to be admitted that no precise explanation can be offered as to why
John should have presumed that entering the praetorium would (or might)
have entailed defilement. All that we know of the period suggests that, with
certain momentary exceptions, the Romans avoided bringing into Jerusalem
images which would offend Jewish sensibilities. Nor is there any reason to
suppose that appearing before the governor would have involved any hospitality by way of food or drink. The context does, however, clearly suggest the
idea either that the location, indoors, itself might impart pollution, or that
contact with gentiles, in a relatively confined space indoors, as opposed to
the open-air courtyard, might do likewise. None the less, however uncertain
we may be as to the precise rules of purity involved, it is not impossible to
Reflections on the Trials of Jesus

find in earlier or contemporary Jewish writing expressions relating to fitness
for Passover which might reflect a general awareness of the need for extra
caution.14 So for instance Ezra :–:
The children of the captivity kept the Pascha upon the fourteenth (day)
of the first month. For the priests and the Levites had purified themselves together; all of them were pure: and they killed the pascha for all
the children of the captivity, and for their brethren the priests, and for
themselves. And the children of Israel, which were come again out of
the captivity, and all such as had separated themselves unto them from
the filthiness of the heathen of the land, to seek the Lord, the God of
Israel, did eat, and kept the feast of unleavened bread seven days with
joy: for the Lord had made them joyful, and had turned the heart of
the king of Assyria unto them, to strengthen their hands in the work
of the house of God, the God of Israel.
The notion of the need for purity as extending to all, not only to priests, is
also expressed very clearly in a source directly contemporary with the crucifixion, Philo’s de specialibus legibus :–:
After the New Moon comes the fourth feast called the Crossing-feast,
which the Hebrews in their native tongue call Pascha. In this festival
many myriads of victims from noon till eventide are offered by the
whole people, old and young alike, raised for that particular day to the
dignity of priesthood. For at other times the priests according to ordinances of the law carry out both the public sacrifices and those offered
by private individuals. But on this occasion the whole nation performs
the sacred rites and acts as priest with pure hands and complete immunity. The reason for this is as follows: the festival is a reminder and
thank offering for that migration from Egypt . . .
Moreover, as J. A. T. Robinson pointed out,15 the Mishnah seems to offer
a conception of a possible context for the incurring, or non-incurring, of
impurity which fits precisely with the presuppositions of John . For the
tractate Oholoth states categorically (:) that ‘‘the dwelling places of gentiles are unclean.’’ We can reasonably conclude that the notions embodied in
John’s narrative are at least not provably inapplicable to this period.
However, far more significance than that may attach to the much discussed
exchange between Pilate and the Jewish leaders, in which he says to them,
. The quotations are borrowed from Segal (n. ),  and .
. Robinson (n. ), –.

Rome and the East
‘‘Take him yourselves, and judge him according to your own law’’ to which
they reply ‘‘It is not permitted to us to execute anyone’’ (:). Their reply
has sometimes been read as an allusion to a fixed and universal ban on the
carrying out of executions (and capital trials?) by the local Jewish authorities, in view of the equally established reservation of that right to the Roman
governor.16 Indeed it has often been quoted as one of the conclusive items of
evidence for the existence of such a rule. If so, however, it must be regarded as
reading very strangely. For the narrative must represent the Roman praefectus
as being unaware of this rule, and as being informed of it by the high-priestly
group before him. Such a reading, however strange, is not however impossible, if we conceive of the exchange as a feature of John’s narrative style,
in which necessary explanations are sometimes given by speakers to respondents who in the ‘‘real world’’ might be presumed not to have needed them.
An example already referred to, occurs a few lines later, when Pilate says,
‘‘You have a custom by which I should release one man to you at Passover’’
(:).
None the less, seeing the exchange as such an authorial device is not the
most natural way of reading the passage; and it does have to be emphasised
that what we are doing is reading a narrative; so how we understand what
we read ought to be determined in the first instance by the nature of the
information and interpretation which the author has already provided. In
that light the most significant guidance and explanation provided by John is
given only four lines before: that it was morning, and that they would not
enter the praetorium because they wanted to avoid pollution and be able to
eat the pascha. In that light the exchange reads quite naturally. Pilate tells
them to judge him (he does not here say execute him) according to their own
law. And they reply that it is not allowed to them to execute anyone. Not
allowed by the Romans? Such an interpretation is possible, as we have seen,
but very strained. Or not allowed by Jewish law? It immediately makes sense,
for we are in the morning before Passover and an execution was surely not
permitted. We hardly need the Mishnah (Sanhedrin :) to tell us, as we have
seen, that capital trials could not be conducted on the day before a Sabbath
or a festival, because a capital sentence could not be pronounced until the
day following the trial. It should be stressed that to emphasise the possibility
of reading the text in this way is not at this point to make any assertion about
‘‘what actually happened,’’ or about the rules of capital jurisdiction which
generally prevailed in the real world of first-century Judaea. It is to suggest a
way of understanding what story John is telling—one to whose entire logic,
. See, e.g., Sherwin-White (n. ), –.
Reflections on the Trials of Jesus

as we have seen, the approach of Passover is fundamental. John reminds us
of this at the moment when Pilate takes his seat on his tribunal: ‘‘And it was
the preparation of the pascha, and it was about the sixth hour’’ (:). So
he does again immediately after Jesus’ death, when the Ioudaioi, since it was
‘‘preparation,’’ ask that the bodies of Jesus and the two robbers may be taken
down so as not to remain on their crosses ‘‘on the Sabbath,’’ by which he
perhaps means ‘‘on the day of a festival’’ (:). If that seems doubtful, we
should note that some lines later, when Mary Magdalen comes early in the
morning, when it is still dark ‘‘on the first (day) of the Sabbath’’ (:), John
surely means ‘‘on the first of the festival days.’’ 17
The structure of the narrative and its religious meaning and context are
derived from its location on the day of preparation and the approach of Passover. It is surely suggestive that one reader who without hesitation ‘‘read’’
:, ‘‘it is not permitted to us,’’ as an allusion to the restrictions imposed
by Jewish law was Saint Augustine: ‘‘But it should be interpreted as meaning that it was because of the sanctity of the festival, which they had already
started celebrating, that they said that they were not allowed to execute anyone; because of this they were afraid of becoming polluted by entering the
Praetorium.’’ 18
Conclusions
This discussion has not attempted to present firm conclusions, which are unattainable, but a series of approaches to the question of how we should attempt to understand our evidence on how Jesus came to be crucified. The
primary suggestion is that in studying both the course of Jesus’ life and the
manner of his death, we must not proceed by amalgamating data from all four
Gospels. That is illegitimate, because not merely the details, but the entire
structure of the story as told by John is different from that in the Synoptics.
Our evidence consists of coherent texts, and even where they contain common items, derived either from each other or from a hypothetical common
source, or sources, any approach must respect the integrity of these texts as
. The interpretation of all these expressions seems to me extremely problematic, and
I am not convinced that any of them clearly indicate what day of the week (as opposed to
what day before, or of, Passover) John means to represent. In particular the Anchor Bible
translation of : as ‘‘the first day of the week’’ seems to me quite impossible. For in that
case Shabbat, as the last day of the week, has disappeared.
. The quotation comes from Augustine, In Joh. Ev. Tract. ,  (CCL XXXVI, ).
The same point is made very briefly by John Chrysostom, In Joh. Homil. ,  (PG LIX,
col. ).

Rome and the East
embodying different narrative structures. Any attempt to answer the inescapable question of ‘‘what really happened’’ must therefore involve a choice.
No arguments for any particular choice can be conclusive, but without such
a choice our selection of elements to prefer must remain merely arbitrary.
Given the necessity of choice, this paper offers the suggestion that, both as
regards the narrative of Jesus’ life and the culminating story of how he met
his death, we should give our preference to John.
The expression of such a preference can in itself be no more than a hypothesis. That is to say that our position should be as follows: if, hypothetically, we accept John’s Gospel as offering us the best account which we have
of the steps which led to the crucifixion, what are the consequences?
Firstly, the arrest, successive examinations, and crucifixion of Jesus took
place not on the first day of Passover, but on the day before, from evening
to midday. The Last Supper was therefore not a paschal meal at which the
paschal lamb was eaten, following the custom by which, by extension, a ceremony originally conducted solely in public in the Temple had become also a
domestic ritual. Instead it was merely a meal on the night before. We would
thus have to accept that it is the Synoptic accounts which have turned it, not
very convincingly, into a paschal meal.19
Secondly, we would have to accept the assertion, unique to John, that the
arrest of Jesus was carried out by a Roman cohort under a tribune, guided
by Judas and assisted by attendants sent by the high priests and Pharisees. If
so, that places him closer to the category of the long succession of popular religious leaders, all viewed as instigators of popular disorder, who are
known from the pages of Josephus, and all of whom, with their followers,
Jewish and Samaritan, were repressed by Roman forces. Jesus must by implication have been viewed as being more like these than like the solitary
and apparently unbalanced pseudo-prophet, the other Jesus, arrested by the
Jewish magistrates at Tabernacles of ..  (see above).
Most important of all, however, is the fact that John’s narrative, in which
Passover has not yet arrived, gives Passover a much more fundamental relevance to what happened—and how—than do the Synoptics, which, while
describing these events as occurring on the first night and morning of Passover, ignore the significance which we must presume to have attached to
it in real life. For it is indubitable that Passover was the most important of
the annual Jewish festivals in this period. Matthew and Mark, it is true, do
not claim any more than that the Jewish authorities arrested Jesus on that
night, examined him in the house of a high priest, and accused him before
. So, e.g., Segal (n. ), .
Reflections on the Trials of Jesus

the Roman praefectus in the morning, pressing on Pilate the necessity of crucifixion, to which he assented, and which then took place. Even that may
seem incredible in view of the requirements of purity imposed during the
festival. But Luke goes further, and represents the calling of a regular council
in the morning, after the examination in the high-priestly house, and before the accusation before Pilate. If such an event really occurred, it must
have offended even more profoundly against the rules later propounded in
the Mishnah—and more importantly against the underlying beliefs about
the sanctity of the festival which gave rise to those rules.
In John, by contrast, Passover, which has not yet arrived, dominates everything. It is because Jesus is brought before Pilate on the morning of the day
whose evening will see the onset of Passover that Jesus’ accusers will not
enter the praetorium. It is, on this interpretation, because the sanctity of the
festival prevents the holding of a capital trial even on the day before, that
they tell Pilate, ‘‘It is not permitted to us to kill anyone.’’ It is because they
(unlike Jesus, who has no choice) remain outside the praetorium, that Pilate
conducts the examination by alternating between questioning Jesus inside
and confronting his accusers, and the crowd, outside, in the paved courtyard. This courtyard, identified only by John, has a regular tribunal (bēma) on
which Pilate formally takes his seat when he finally brings Jesus outside to
confront the Jewish crowd. Nothing is described here, however, any more
than in the other Gospels, which could count as a formal trial by Pilate. There
is no formal accusation and defence; no opinions are asked of the governor’s
council; and no formal verdict is pronounced. As with the other Gospels,
in John the decision by Pilate to have Jesus executed is not represented as a
verdict concluding a trial, but as a political decision taken as a concession to
political pressure both from the Jewish authorities and from the crowd. The
long-debated question of whether it was the Roman governor or ‘‘the Sanhedrin’’ which had the legal right to try capital cases in first-century Judaea
and to order the execution of those condemned on capital charges may be
doubly misconceived, if it is thought to be directly relevant to how we interpret the Gospels. For, firstly, it has already been suggested that we should
think rather of the pattern presented by Josephus’ account of the execution
of James, when the high priest summons ‘‘a council [synedrion] of judges’’
(Ant. , ); that is to say, when the occasion arose, the high priest called
together a group of citizens of his own choosing, just as a Roman official
would summon a council (consilium).20 Secondly, of the four Gospels, only
. For this suggestion, see M. Goodman, The Ruling Class of Judaea: The Origins of the
Jewish Revolt against Rome (), –.

Rome and the East
Luke represents anything that we could think of as a meeting of ‘‘the Sanhedrin’’ and does so in a context (the first day of Passover) which is highly
improbable. For the rest, what we see is an examination in the house of a
high priest; or rather, in John’s case, of the two successive high priests, whose
current roles are carefully (and correctly) specified; again the distinction between the high priest currently in office (Caiaphas) and an ex-holder with
the rank of high priest is unique to him. Since no Gospel represents Pilate’s
decision as a formal verdict, there is a very clear sense in which the entire
notion of ‘‘the trial of Jesus’’ is a modern construct.
If instead it was a political decision, then it is again John who gives
the clearest conception of its motivation. As Martin Goodman points out,
Annas/Ananus was the first high priest to be appointed by a Roman praefectus
after the provincialisation of Judaea in .. , and he and Caiaphas between
them occupied the position for all but about three of the first twenty years
of the provincial period. They cannot but be seen as having collaborated successfully with the occupying power. That Caiaphas actually expressed the
view that it was worth sacrificing one man to prevent Roman retribution on
the whole people ( Jn. :) cannot of course be known. But the thought
would have corresponded well enough to the logic of the political situation.
Moreover, either we know nothing whatsoever about the crucifixion or we
can at least accept the one detail on which all four Gospels agree, that on the
cross Jesus was described as ‘‘King of the Jews.’’ Only John adds the detail that
this inscription was called by the borrowed Latin term titlos (titulus), and that
it was written in three languages. More important, it is he who represents
Pilate’s final dialogue with the crowd as turning on just this point: ‘‘If you
release this man, you are not a friend of Caesar; for anyone who makes himself a king opposes Caesar,’’ ‘‘Behold your king!’’ ‘‘Take him and crucify him,’’
‘‘Shall I crucify your king?’’ ‘‘We have no king but Caesar.’’ Philo’s Against
Flaccus happens to provide a precisely contemporary analysis of the susceptibility to such pressure of a Roman governor who feels himself to be out of
favour with the emperor (Flacc. –/–).
It was still the day of preparation for Passover ( paraskeuē ), about the sixth
hour; there was still time for the crucifixion to take place, and for the bodies
of Jesus and the two robbers to be taken down before the festival proper
began (:), though the tomb where the body was laid had to be close because of the ‘‘preparation of the Jews.’’ By the next morning it was already of
course the first day of the festival, or, as John puts it, ‘‘the first day of the Sabbath’’ (:–:). It is the approach of Passover which dictates every aspect
of how the story unfolds, just as, in John’s narrative, Jesus’ life as a Galilean
holy man is structured by the need to go up repeatedly to Jerusalem to the
Reflections on the Trials of Jesus

annual cycle of festivals. This necessity should, I suggest, be seen as of crucial
importance. It is remarkable that, even in recent years, it has seemed possible
to discuss the ‘‘Palestinian Judaism’’ of this time as a purely personal religion,
without giving a central (or indeed any) place to the communal worship and
sacrifice at the Temple.21 But John’s narrative may precisely reveal the centrality of this communal sacrificial cult in the life of Jesus. If his portrayal
gives us a Jesus who is less far from ‘‘the historical Jesus’’ than the one whom
the Synoptics represent, then we can at least perceive that their attachment
to these festivals was one thing which Jesus and his accusers shared. It was
not Roman law but their own which made them say, at that moment, ‘‘It is
not allowed to us to execute anyone.’’
. So E. P. Sanders, Paul and Palestinian Judaism (), part I: ‘‘Palestinian Judaism.’’
 
The Roman Coloniae of the Near East:
A Study of Cultural Relations *
Introduction
‘‘The history of Roman colonisation is the history of the Roman state’’: so
Ernst Kornemann, in his standard article on Roman coloniae.1 The following survey of the coloniae which the emperors created between the late first
century .. and the middle of the third century .. in the Fertile Crescent—or, on a different definition, in those provinces of the Roman Empire
where both Greek and various Semitic languages were spoken—cannot of
course contribute directly to an understanding of the earlier phases of that
history. But it does very vividly reflect the rapid progression in the nature
of the ‘‘colonising’’ process itself, from the one unquestionable military colonia of the Augustan period, Berytus, with its wide territory extending over
the Mount Lebanon chain into the Bekaa valley, to the three coloniae of the
first and second centuries .., all in or on the borders of Judaea, to a wholly
new phase in the Severan period and the following decades, when the title
colonia was granted to towns all over the region, including the new province
of Mesopotamia, and thus came into use almost as far east as the Tigris.
* First published in H. Solin and M. Kajava, eds., Roman Eastern Policy and Other Studies in
Roman History (Helsinki, ), –.
This paper, a Vorarbeit for The Roman Near East ( ..–.. ) (), owed an immense amount to comments, corrections, and additions from various friends and colleagues, Sebastian Brock, Hazel Dodge, Louis H. Feldman, Martin Goodman, Christopher
Howgego, Beniamin Isaac, Nikos Kokkinos, Barbara Levick, Aharon Oppenheimer, JeanPaul Rey-Coquais, and Alla Stein. I was also very grateful to Heikki Solin and his colleagues
in Helsinki for the colloquium in autumn  which provided the impulse for the paper,
and for contributions on that very enjoyable occasion, and to Mika Kajava for the care taken
in the editing of this complex text.
. F. Kornemann, RE IV (), s.v. ‘‘coloniae,’’ cols. –, in col. ; ‘‘Die Geschichte
der römischen Kolonisation ist die Geschichte des römischen Staates.’’

The Roman Coloniae of the Near East

But the process of colonisation, or of the conferment of the title colonia,
was also one of many forms of intervention by Rome in the social structures and communal identities of a region long since Hellenised, but where
a variety of other ethnic identities, most of them extremely difficult for us
to characterise without distortion, were still very important factors.2 Roman
colonisation naturally introduced a new element into this already complex
scene, and in this sense it can be perceived from a quite different angle, not
as part of the history of Rome but as an element in the cultural history of the
Fertile Crescent. Of all the Roman coloniae of this region, only one, Berytus,
with its hinterland which was later separated off as the colonia of Heliopolis,3
represented a substantial island of Romanisation, of Latin language and culture, and of Roman law, which was to last into the late Empire. But this
colonia too was established in the context of an already-existing Greek, or
Graeco-Phoenician, city, and inevitably took on many of the roles and forms
of public life associated with a Greek city. The same was a fortiori true of the
others, all of which, with one exception, show a profound continuity with
the Greek cities which preceded them. The exception is Aelia Capitolina,
founded by Hadrian on the ruins of Jerusalem, destroyed as a Jewish city in
the great revolt of .. , and now recreated as a pagan city after the Bar
Kochba war of .. –. But even Aelia, like the other coloniae of this region, and unlike almost all those of Italy, Africa, and Europe, continued to
mint coins until the mid-third century.4 For minting was a normal, if not
universal, characteristic of cities in this region, and nearly every colonia, real
or titular, acted likewise. The rank of colonia was to become, among other
things, simply another status, like that of metropolis, to which a city might
aspire. In contemporary literature or documents, depending on the context,
the title colonia might deserve mention, or might not. The latter, for instance,
was the case when a boxer put up an inscription at Aphrodisias in Caria in
about ..  to record his world-wide victories. When he lists the places
in the Syrian region where his victories had been won, his words represent
a wholly Greek world, in which Roman colonisation is invisible, just as is
. For a sketch of some of the issues, see F. Millar, ‘‘Empire, Community and Culture
in the Roman Near East: Greeks, Syrians, Jews and Arabs,’’ Journal of Jewish Studies  ():
; see also F. Millar, The Roman Near East ().
. This formulation presupposes the validity of a particular view of the earlier history
and status of Heliopolis, discussed more fully below.
. The coloniae of Spain, Africa, and Gaul ceased to mint, as did other towns in these
regions, in the Julio-Claudian period. The one remarkable exception to the rule that coloniae in the Latin-speaking provinces did not mint after the mid-first century is provided
by Viminacium, which minted between the reigns of Gordian and Gallienus.

Rome and the East
the non-Greek element: ‘‘at Damascus, the men’s pancration [all-in wrestling]
twice, at Berytus the men’s pancration, at Tyre the men’s pancration, at Caesarea
Stratonos the men’s pancration, at Neapolis of Samaria the men’s pancration,
at Scythopolis the men’s pancration, at Gaza the men’s pancration, at Caesarea
Panias the men’s pancration twice, at Hieropolis the men’s pancration. . . .’’ 5
Berytus and Caesarea were already coloniae (and, as it happens, Damascus,
Tyre, Neapolis, and Gaza were later to gain the same status).
That was not, however, the whole story. The impact of Roman rule was
reflected, among other things, by the entry not only into Greek but into
Semitic languages of a long list of Latin terms, among them colonia and its
cognates.6 Some of these will be considered further below, above all in connection with Palmyra and Edessa. But it may be instructive to mention here,
as an example of the complex cultural context into which these Roman coloniae, or nominal coloniae, were inserted, two inscriptions, one from Apamea
and one from Palmyra. In Apamea, in the reign of Hadrian, we find the peculiar institutions of the colonia of Berytus reflected in an inscription honouring
an actor: among other distinctions he was ἐν κολωνείᾳ Βηρύτῳ τετειμημένον σεξβερὰτι (AE , ). The term is not other-wise attested in Greek
transliteration, and the form in which it is given is presumably intended to
represent the Latin ablative sexviratu. Then, from a century later, there is a
Palmyrene bust of the first half of the third century, now in the Louvre,
which represents a man with a Greek-Latin name, who is a citizen not of
Palmyra itself but of the colonia of Berytus.7 The subject is identified in a
. L. Moretti, Iscrizioni agonistiche greche (), no. .
. For Greek, see H. J. Mason, Greek Terms for Roman Institutions: A Lexicon and Analysis
(),  and –. For Semitic inscriptions, see M. G. Angeli Bertinelli, Nomenclatura
pubblica e sacra di Roma nelle epigrafi semitiche (),  (colonia, colonus); – (duumvir,
duumviratus). In the latter case the relevant transliterated forms are all borrowed from the
Greek equivalents. See, for comparative material, S. Krauss, Griechische und lateinische Lehnwörter im Talmud, Midrasch und Targumim (–), and A. Schall, Studien über griechische
Fremdwörter im Syrischen (). For QLNY’, see M. Jastrow, A Dictionary of the Targumim
(repr. ), . For linguistic interplay, note also M. G. Angeli Bertinelli, ‘‘I Semiti e
Roma: appunti di una lettura di fonti semitiche,’’ Serta Historica Antiqua I (), . Martin
Goodman drew my attention to further uses of colonia as a loan-word in the Babylonian
Talmud: Sukkah a; Yebamoth b; Baba Bathra a. QLNY’ unfortunately does not find a
place in D. Sperber, A Dictionary of Latin and Greek Legal Terms in Rabbinic Literature ().
. C. Clermont-Ganneau, Répertoire d’épigraphie semitique II, no.  CIS II ;
(Greek) IGR III  OGIS . See M. A. R. Colledge, The Art of Palmyra (), 
and pl. .
The Roman Coloniae of the Near East

bilingual inscription, in Greek on the left and Palmyrene on the right, as
follows:
Μᾶρκος Ἰούλιος / Μάξιμος / Ἀριστείδης, / κόλων / Βηρύτιος, /
πατὴρ Λου/κίλλης γυ/ναικὸς Περ/τίνακος
MRQWS YWLYWS MKSMWS / ’RSṬYDS QWLWN / BRTY’
’B(W)H DY / LWQL’ ’TT PRṬNKS
The Latin word colonus has thus become a pseudo-Greek word, κόλων, and
from there has been transliterated into Palmyrene as an intelligible form of
identification of a person from the colonia of Berytus. Similar transliterated
forms were already in use for the institutions of Palmyra itself, or very soon
would be, as we will see (text to nn. –). But, as we will also see, allusions to cities as coloniae and to their inhabitants as preserving the relevant
status are very rare after the third century; it remains unclear whether this is
because such statuses had genuinely ceased to be relevant; or because the city
coinages of the Greek East had ceased in the second half of the third century,8 or because the ‘‘epigraphic habit,’’ though it had not ceased, had greatly
declined. Moreover, Palmyrene, a language which is known only through inscriptions, and whose vocabulary is exceptionally rich in loan-words, ceased
to be inscribed soon after the reconquest by Aurelian.9
It is therefore all the more striking that even in the late fifth century, in a
period when to all appearances the status of colonia had long ceased to be of
any importance, we still find preserved in the Babylonian Talmud a reflection of the possibility which had existed in the early Empire, of asking the
emperor to grant the rank of colonia to a city. For the treatise Avodah Zara (on
pagan worship) contains a tale of an emperor called ‘‘Antoninus’’ (conceivably Caracalla) saying to ‘‘Rabbi’’ that he wishes ‘‘to make Tiberias a colonia’’—
TT‘BYD ṬBRY’ QLNY’.10 This elevation, so far as we know, did not in fact
. See the useful presentation of these coinages, and their disappearance, by K. Harl,
Civic Coins and Civic Politics in the Roman East, .. – ().
. For this point, see Millar (n. ), . Note, however, the two Palmyrene inscriptions
published and discussed by M. Gawlikowski, Syria  (): –, and Le temple palmyrénien (), , nos. –, of which one dates to , while its companion (of ?) seems
to refer to Aurelian—[’W]RLYNWS QSR. Note also the pair of Greek and Palmyrene inscriptions of .. /, M. Gawlikowski, Ann. Arch. Ar. Syr. – (–): , no. .
. Bab. Talmud, Avodah Zara, a. Note that in Tiberias: From the Foundation to the Moslem
Conquest (), – (Hebrew), Y. Meshorer claims to read COL on coins of the city under
Elagabal. I owe this reference to Alla Stein.

Rome and the East
occur, and the story may very likely be legend. But the Severan period was
exactly that in which a whole series of places not far away—Heliopolis, Tyre,
Sebaste, Emesa, Sidon, Petra, Bostra, followed in the midcentury by Neapolis and Philippopolis—did gain this status. It had been something to which
a middle-rank town like Tiberias could then reasonably have aspired.
That aspiration, whether a historical reality in itself or not, reflects the
third, final, and most complex stage of Roman ‘‘colonisation’’ in the Near
East. The three stages can be regarded as quite distinct: firstly, the settlement
of a veteran colonia at Berytus under Augustus; secondly, the three coloniae of,
or on the borders of, Judaea, namely Ptolemais, Caesarea, and Aelia Capitolina, founded between the mid-first and mid-second century ..; and,
thirdly, the widespread grants of the status of colonia in the Severan period
and the mid-third century, extending far into the newly conquered territory
of Mesopotamia. The successive phases will be discussed in order here; given
the importance and novelty of the nature of ‘‘colonisation’’ under Severus,
that period will be discussed separately from the remaining period up to the
mid-third century.
Berytus and Its territorium; Heliopolis
Alone of all the coloniae to be considered here, Berytus belongs in the only
major phase of organized veteran settlement outside Italy in Roman history,
the age of Caesar and Augustus.11 Before that period organised colonial settlement outside Italy was almost unknown: the colonia of Narbo in southern
Gaul ( ..) was indeed the only example of a pre-Caesarian provincial
colony which was successfully established and survived as a formal entity.
After the reign of Augustus, actual coloniae, involving the establishment of
settlers and the formation of a new city constitution, can be found; but they
are not common, and it becomes increasingly difficult to distinguish such a
settlement from the mere conferment of the title colonia on an existing city,
with some or all of the associated rights. An extra complication is added to
the question by the fact that one of our best and most used items of evidence
on the formal and legal character of coloniae comes not only from the main
. Among more recent works on this major topic, note F. Vittinghoff, Römische Kolonisation und Bürgerrechtspolitik unter Caesar und Augustus (); B. M. Levick, Roman Colonies
in Southern Asia Minor (); P. A. Brunt, Italian Manpower,  ..–..  (), chap. 
and app. ; L. J. F. Keppie, Colonisation and Veteran Settlement in Italy, – .. (); J. C.
Mann, Roman Colonisation and Veteran Settlement during the Principate ().
The Roman Coloniae of the Near East

period of the creation of nominal coloniae, but from a citizen of one of them,
Domitius Ulpianus from Tyre (text to n.  and following it).
Berytus, however, represents a quite different case.12 Strabo’s contemporary account of its settlement, though not entirely clear as to the precise
extent of its territorium, provides all the essentials: ‘‘But though Berytus was
razed to the ground by Tryphon, it has now been restored by the Romans;
and it received two legions, which were settled there by Agrippa, who also
added to it much of the territory of Marsyas as far as the Orontes river.’’ 13 Of
the history of Berytus as a Hellenised Phoenician city in the late Hellenistic and then Roman republican periods very little can be known; but Strabo
records, just before the passage quoted above, that robbers established on
Mount Lebanon had been able to ravage both Berytus and Byblos (, , 
[]); and the coinage of Berytus indicates that the city had been included
in Antonius’ grant of the coast of Phoenicia to Cleopatra, made perhaps in
/ ..14 The evidence is sufficient to make clear that Berytus will have
shared in the generally disturbed history of the region in the mid-first century ..,15 without providing any conclusive reason why the one Roman
colonia should have been placed there rather than anywhere else. But Strabo’s
words suggest a military purpose, in relation to the unsubdued mountain
zone behind; or rather two mountain zones, the chain of Mount Lebanon
and to the east that of Anti-Lebanon, with the Marsyas, or Bekaa valley, and
the sources of the Orontes, in between.
Jerome’s Chronicle offers a firm date for the foundation of the colonia, as of
the colonia of Patras,  ..16 Since Strabo (above) attributes the foundation
to Agrippa, and this was a period when he was present in the Syrian region,
the dating can be accepted. It is not impossible, however, in the case of Patras
. For the history of the city, see R. Mouterde and J. Lauffray, Beyrouth ville romaine: histoire et monuments (); R. Mouterde, ‘‘Regards sur Beyrouth phénicienne, hellénistique et
romaine,’’ Mél. Univ. St. Joseph  (): ; N. Jidejian, Beirut through the Ages (); J. Lauffray, ‘‘Beyrouth Archéologie et Histoire, époques gréco-romaines I. Période hellénistique
et Haut-Empire romain,’’ ANRW II. (), . For Berytus as a colonia, see B. Isaac, The
Limits of Empire: The Roman Army in the East (), –, a study to which this paper
owes much.
. Geog. , ,  (), Loeb trans.
. For the details, see Schürer, Vermes, and Millar, History I (), –.
. For an excellent survey of the history of the region, to which what follows owes
much, see J.-P. Rey-Coquais, ‘‘Syrie romaine, de Pompée à Dioclétien,’’ JRS  (): .
. R. Helm, Die Chronik des Hieronymus (), : Coloniae Berytum (sic) et Patras deductae. Bosforum Agrippa capit.

Rome and the East
that a settlement of legionaries after Actium was followed only later by the
formal creation of a new colonia, which incorporated a substantial number
of inhabitants of the surrounding region.17 If that sequence is valid, which
is not wholly certain, the same may perhaps have applied to Berytus. But if
some slight uncertainty prevails as to the date, the essential fact that this was
a genuine veteran colonia is secure. It is consonant with that fact that both
its foundation by the favour of Augustus continued to be recalled, and that
Berytus was regarded as possessing what was later at least described as the ius
Italicum. So Ulpian writes in the first book of his de censibus: Sed et Berytensis
colonia in eadem provincia (Syria Phoenice, as it had become in Ulpian’s time),
Augusti beneficiis gratiosa, et (ut divus Hadrianus in quadam oratione ait) Augustana
colonia, quae ius Italicum habet (Dig. , , , ). Ulpian’s quite flowery Latin
is not free from obscurity; but in the context he clearly means (at least) that
the new colonia paid no tributum.
Effectively nothing is known of the urban character of late Hellenistic
Berytus, and we do not know whether or not a fundamental restructuring
of the urban plan will have taken place in the case of a colonia established
on the site of an existing city. The ancient rituals for the foundation of a
colonia are indeed represented, along with the signa of the legions V Macedonica and VIII Augusta, on coins issued by Berytus from the reign of Augustus onwards.18 But coins representing a founder ploughing with an ox and
a cow, while they may be records of an actual historical event, the marking out of a new boundary for Berytus in  .., cannot be assumed to be
so. Their significance may be purely symbolic. But, while we also do not
have any evidence for an urban restructuring of Berytus to compare with
Josephus’ account of Herod’s re-foundation of Stratōnos Pyrgos as the new
city of Caesarea,19 we do know that Herod’s munificence in providing public buildings for cities in the Syrian region embraced Berytus also: ‘‘Thus he
provided gymnasia for Tripolis, Damascus and Ptolemais, a wall for Byblos,
. See J.-M. Roddaz, ‘‘Marcus Agrippa,’’ BEFAR  (): –, depending, however, on a combination of Strabo , ,  (), referring to a settlement of legionaries subsequent to Actium, and Pausanias , , , describing the establishment of a new colonia with
a civilian population from the surrounding region. No certain chronological sequence can
be deduced. For arguments for an early date for the settlement of at least some veterans in
each place, see J. G. P. Best, ‘‘Colonia Iulia Equestris and Legio Decima Equestris,’’ Talanta
 (): .
. G. F. Hill, BMC Phoenicia (), lv. f., e.g., no.  and pl. VIII.. See R. Mouterde,
MUSJ  (): , on –.
. Josephus, BJ , –; Ant. , –. On Caesarea, see below.
The Roman Coloniae of the Near East

halls (exedras), porticoes, temples and market-places (agoras) for Berytus and
Tyre, theatres for Sidon and Damascus, an aqueduct for Laodicea on the sea,
baths, sumptuous fountains and colonnades, admirable alike for their architecture and their proportions, for Ascalon.’’ 20 It is interesting that Josephus,
although of course aware that Berytus was a Roman colonia,21 does not see
Herod’s contribution to public building there as different in kind from that
which he made in the case of Greek cities.
The evidence does not in fact allow us to say more than that Berytus, in the
imperial period, acquired the normal apparatus of public buildings, adorned
with Corinthian columns, which characterised the more important provincial cities. The fragmentary archaeological and architectural traces recovered
from beneath the modern city confirm this, without proving that the city
was distinctively Roman in appearance, or different from the surrounding
Greek cities.22
The munificence of Herod was followed by that of his grandson Agrippa I.
But here perhaps we can see reflected the particular status and particular cultural character of the place. For Agrippa, as well as building a theatre, provided an amphitheatre, something not unknown, but certainly not typical,
in Greek cities,23 and then laid on there gladiatorial shows involving ,
condemned criminals.24 The gladiatorial show was one of the most distinctive Roman importations into the communal life of the Greek East, and
this report represents one of the earliest items of evidence for its intrusion
there;25 it is perhaps not an accident that the shows took place, in a newly
built amphitheatre, in a Roman colonia.
We cannot be equally sure that there was anything distinctively Roman
about the benefactions of Agrippa’s son, Agrippa II, in the reign of Nero.
He too is recorded as having built a theatre (a new one, or rebuilding the
existing one?), and having put on lavish shows; but he also distributed both
grain and oil, the latter surely more characteristic of popular expectations
in a Greek city. Beyond that he adorned the city with statues and ‘‘eikones
. Josephus, BJ , . Loeb trans. Not repeated in Ant.
. BJ , : ἡ δ᾿ ἐστὶν ἐν τῇ Φοινίκῃ πόλις Ῥωμαίων ἄποικος.
. For the literary and archaeological evidence for Roman Berytus, see, above all, Mouterde and Lauffray (n. ).
. See, e.g., F. Millar, ‘‘Introduction: The Greek World and Rome,’’ in S. Macready and
F. H. Thompson, Roman Architecture in the Greek East (). Hazel Dodge, however, pointed
out to me that recent discoveries had revealed more amphitheatres than previously known.
. Josephus, Ant. , –.
. See L. Robert, Les gladiateurs dans l’Orient grec (; repr. ), –.

Rome and the East
which were copies of ancient ones.’’ 26 It is far from clear what this means.
Were they copies of classical Greek sculptures (like the statue of Augustus in
his temple at Caesarea, modelled on Pheidias’ Zeus at Olympia, and the Rōmē
modelled on the Hera from Argos)? 27 Or something like the bronze statues
of figures from Greek mythology, installed in the early second century at the
new baths in Apamea on the Orontes? 28 Or were these images ones which
more specifically recalled the Roman character and origins of the city?
The latter is not impossible. For, while there is every reason to presume
that the new colonia incorporated the existing inhabitants of Berytus and its
hinterland as far as Anti-Lebanon, and while it is generally characteristic of
those Roman coloniae in the Greek East whose epigraphy has been studied
that the use of the Greek language steadily reasserted itself,29 it is very distinctive of Berytus that its Roman and Latin character was heavily emphasised from the beginning and remained clearly visible at least until the fourth
century. The city’s public use of Latin is for instance still embodied in an inscription of .. , from the base of a statue at Berytus representing Flavius
Domitius Leontius, praetorian prefect and consul ordinarius: ordo Berytiorum
statuam sumptibus suis ex aere locatam civili habitu dedicavit.30 Without unnecessarily multiplying examples, the public use of Latin there can easily be traced
back through the Tetrarchic (CIL III 7 ) and Severan 31 periods to the
second century. It was then, as it seems, that for instance M. Sentius Proculus was a decurion and duumvir, and subsequently embarked on an equestrian
career before entering the Senate, and being honoured as patronus coloniae
with at least two statues, as their bases, inscribed in Latin, attest.32 From the
second half of the first century a Latin inscription records the restoration by
Agrippa II and Berenice of a building whose name is lost [q]uod rex Herodes,
proavos eorum fecerat, adorned with marble and columns (AE , ).
Yet another Latin inscription comes from an altar dedicated to the Fortuna of the Genius of the Colonia by M. Iulius Avidius from Emesa, who
. Josephus, Ant. , –: καὶ τὴν πᾶσαν δὲ πόλιν ἀνδριάντων ἀναθέσεσιν καὶ ταῖς
τῶν ἀρχαίων ἀποτύποις εἰκόσιν ἐκόσμει.
. Josephus, BJ , .
. AE , .
. For the coloniae in Pisidia, see Levick (n. ), chaps. –; for Corinth, see Corinth
VIII.: The Inscriptions, – (), –.
. ILS .
. See, e.g., Bull. Mus. Beyrouth  (–): , no.  AE , , a dedication to
Iulia Domna.
. See R. Cagnat, ‘‘M. Sentius Proculus de Beyrouth,’’ Syria  (): ; AE , ;
H. Devijver, Pros. milit. Equestr. II, S. .
The Roman Coloniae of the Near East

was honoured with a corona by the decuriones, apparently at the moment of
the centenary of the foundation of the colonia—saeculo c(ondita) c(olonia)—so
(see above) in the s .. (AE , ). If that is so, it had been not long
before that, in the reign of Vespasian, that someone, perhaps the Emperor
himself, had erected or restored tabernae, perhaps fronting the forum, and
had put up a statue of Liber Pater, [s]ignum Liberi Patris.33 Given the association, or confusion, of the two, it may well be that this statue is the same as
that of Liber Pater’s servant Marsyas, which coins of Berytus from the reign
of Elagabal represent as standing in front of an arched entrance which may
be an entrance of the forum. If so, and of course that remains a speculation,
it might have been a deliberate reminiscence of the statue of Marsyas which
stood on the edge of the comitium in Rome.34 Marsyas also appears on some
of the earlier coins of the colonia. Here as elsewhere, however, the question
of what exactly that signified remains open to debate. There is no concrete
evidence whatsoever for the common notion that Marsyas specifically denoted the possession of ius ltalicum. Servius (ad Aen. , ; cf. , ) believed
rather that a statue of Marsyas placed in the forum of a city had indicated
that it enjoyed libertas. Veyne argued that Marsyas on city coinages showed
no more than a claim to association with Rome; but a clear correlation with
colonial status is indeed evident. Beyond that, nothing is certain.35
Given the relatively restricted corpus of inscriptions from Berytus, never
collected and hardly likely to be in the foreseeable future, the attestation
there of the cults of Roman deities in private dedications (though not on the
city coins) is extraordinarily full: Venus Domina and Mercurius Dominus;36
Venus, Mercurius, Apollo, Diana, Mars, Ceres(?), Proserpina in a single dedication, along with the Fortuna of the Colonia and the Fata;37 and above all
the ancient and little-known deity, Mater Matuta, whose temple stood on
. CIL IV , subsequently re-published; see R. Mouterde, ‘‘Monuments et inscriptions de Syrie et du Liban . L’emplacement du Forum de Béryte,’’ MUSJ  (–): ,
whence the following identifications are derived.
. See M. Torelli, Typology and Structure of Roman Historical Reliefs (), ; F. Coarelli,
Il Foro romano II: Periodo repubblicano e augusteo (), –.
. For Marsyas on coins of Berytus, see BMC Phoenicia, –. See P. Veyne, ‘‘Le Marsyas ‘colonial’ et l’indépendence des cités,’’ Rev. Phil.  (): . Cf. F. W. Klimowsky,
‘‘The Origin and Meaning of Marsyas in the Greek Imperial Coinage,’’ Isr. Num. Journ. –
 (–): ; and for a reassertion of an original connection with colonial foundations,
Torelli (n. ), –. For ius Italicum, note also M. Malavolta s.v. in Diz. Epig. IV, fasc. –
(), –.
. Syria  (): , nos. – AE , –.
. Syria  ():  AE , .

Rome and the East
the Forum Boarium in Rome.38 The inscription naming this goddess (CIL III
) comes from a place called Der el-Kala in the hills behind Berytus where
there was a temple of Zeus Baalmarcod, of which more in a moment.39 The
Latin inscription records that an altar was dedicated to Mater Matuta by a
lady named Flavia Nicolais (quae et?) Saddane, ex responso Deae Iunonis. How
a Graeco-Syrian woman, evidently a new citizen, came to subscribe to the
cult of Mater Matuta we can only guess. But for the name of such a deity to
have been familiar at all implies a profound ‘‘Roman’’ consciousness among
at least some of the population. The inscriptions of Berytus and its hinterland are sufficient at least to raise the question (which is a pure speculation)
as to whether the eikones imitating ancient models which Agrippa II installed
might have been intended as a compliment to the colonia, and have been borrowed from Rome rather than classical Greece.
The inscriptions also serve to put into an intelligible context the early
experience of the famous Latin grammarian Valerius Probus, from Berytus:
‘‘In the province he had read with a grammatistes certain ancient libelli, for the
memory of the antiqui still survived there, and had not yet been erased as it
had in Rome.’’ The period concerned is the first half of the first century ..;
it remains equally noteworthy that Probus turned finally to his grammatical studies only after long seeking a centurionate and giving up in despair.40
Had he been successful, he would have been another in the series of centurions and primipili, with the tribe Fabia, from Berytus/Heliopolis, as well as
common soldiers, attested on Latin inscriptions both locally and abroad.41
It was this very distinctive character of Berytus as a place of Roman culture
which was to give it its slowly developing role as the site of a major school
of Roman law. To put it in those terms, however, is to go beyond what our
evidence for the period up to the fourth century actually says. Nothing at all
indeed attests this role until we come to the address of thanks which Gregorius, the later bishop and ‘‘wonder-worker’’ (Thaumatourgos), from Neocaesarea in Pontus, addressed to his teacher Origen, probably in the s. In
. See F. Coarelli, Il Foro Boario dalle origini alla fine della Repubblica (), chap. .
. For the site and the temple, see D. Krencker and W. Zschietzschmann, Römische Tempel in Syrien I (), –, with vol. II, . (map) and .
. Suetonius, de gram. .
. From Heliopolis and its vicinity, e.g., IGLS VI  (L. Antonius Naso, career beginning with centurionate); IGLS VI / (L. Gerellanus Fronto, primuspilus); IGLS VI
 (C. Velius Rufus). Common soldiers: ILS  (Coptos); AE , ; ,  (both
Carnuntum); ,  (Rome). Another is attested in the military documents from Masada,
published by H. M. Cotton and J. Geiger, Masada II: The Latin and Greek Documents (),
no. , to whose commentary I owe these references.
The Roman Coloniae of the Near East

it he explains that the ambition which had first led him to the Syrian region
had been a desire to pursue the study of Roman law, and for that he and his
brother had set out for Berytus, ‘‘a city somewhat more Roman, and considered a place of instruction in those laws.’’ 42 The reference is not to a ‘‘school,’’
but to Berytus itself as a place of Roman culture, including legal culture. By
the end of the century we find Diocletian and Maximian replying to students
from the province of Arabia who are engaged on liberalia studia, and especially the professio iuris, and hence are living in civitate Berytensium provinciae
Phoenices, to say that they may have exemption from munera until age twentyfour (CJ , , ). The view that students from the Greek East could find in
Berytus a ‘‘Roman’’ or Latin culture of which legal studies were a prominent
part is strikingly confirmed by the verse epitaph from the fourth century
of a young man from Colybrassus in Cilicia: ‘‘When I journeyed from here
to the distinguished city of Berytus for the sake of the Roman Muse and
the laws . . .’’ 43
We need not pursue the story further, or repeat the well-known allusions
by Libanius to the attraction of students of Roman law to Berytus.44 Enough
has been said to show that Roman culture was a very distinctive feature of
the city into the fourth century. That is not to say, however, that Berytus was
ever a zone of purely Roman culture. Franz Cumont’s description of Berytus, Heliopolis, and Ptolemais as ‘‘Latin islands in the Semitic ocean’’ 45 was
indeed both overstated and misleading in a variety of different ways. For at
start there is no reason to suppose that the Greek-speaking population of Augustan Berytus was not incorporated in the colonia. As a symbol (if no more)
of that continuity, Poseidon, the chief deity of Hellenistic Berytus, reappears
in the second and third centuries on the coins of the colonia.46 It is also interesting that the Greek writer Hermippus, an antiquarian from Berytus of the
early second century, who wrote among other things a work ‘‘On Slaves Who
Distinguished Themselves in Paideia,’’ is described by the Suda as originating
‘‘from an inland village.’’ 47 For the possession of an extensive hinterland embracing parts of two mountain ranges and the Bekaa valley is an important
. H. Crouzel, Grégoire le Thaumaturge remerciement à Origène (Sources Chrétiennes ,
), /: ἡ τῶν Βηρυτίων πόλις ἡ δὲ οὐ μακρὰν ἀπέχουσα τῶν ἐνταῦθα [Syria Palaestina] πόλις Ῥωμαϊκωτέρα πως, καὶ τῶν νόμων τούτων εἶναι πιστευθεῖσα παιδευτήριον.
. SEG XXVI, , with references to the literature.
. For the later history, see P. Collinet, Histoire de l’école de droit de Béryte ().
. CAH XI 1 (), , quoted by Vittinghoff (n. ), .
. BMC Phoenicia, lxvi–ixi. A dolphin-and-trident type had, however, appeared under
Augustus (no. ).
. Suda, ed. Adler II, , s.v. Ἕρμιππος, Βηρύτιος, ἀπὸ κώμης μεσογαίου.

Rome and the East
feature of what colonial Berytus was. But in fact such a Greek writer might
easily have come from the city itself. As we would expect, though the coins
of the colonia remain (unlike those of some later nominal coloniae; see below)
firmly Latin, as do inscriptions recording votes of the decuriones, private inscriptions from there may be in Greek as well as Latin. The ‘‘sea’’ in which
Berytus was an island cannot be described in any simple way as ‘‘Semitic.’’ If
‘‘Semitic’’ is supposed to have some wider cultural or racial connotation, the
criteria for its use are obscure, and its employment is potentially misleading.
If it has a linguistic sense, which makes it at least susceptible to definition,
then it should be conceded that there is not a single Semitic language inscription from the city or its territory in the colonial period, except for one
stray Palmyrene one from Harbata in the Bekaa.
That is not to say that there were no non-Graeco-Roman elements to be
found in the culture of the region, or no Semitic-language terms embedded
in Greek or Latin. A clear example, which is also of considerable importance
for understanding Heliopolis (see further below), is presented by the sanctuary of Theos Baalmarcod at Der el-Kala, mentioned above. The meaning of the
Semitic name of the god is conveniently supplied by a Greek dedication of
the Roman period, where he is addressed as ‘‘Lord of dances,’’ 48 corresponding clearly to B‘L MRQD. We need not doubt that this is a local cult, taken
over and observed by the new population of the colonial period. There is
nothing to show how long the cult had already been established there; all that
is clear is that the name of the deity is Semitic, and that dedications, in both
Greek and Latin were made there under the Empire.49 Similar dedications
might be made to Baalmarcod in Berytus itself, and the wording of one of
them is of particular interest: I.O.M. Balmarcodi M. Verginius Bassus (centurio)
leg. IIII Scyt. vot. sol.50 The appellation Iuppiter Optimus Maximus Balmarcod is
repeated elsewhere as I.O.M.B.51
When the Latin-speaking inhabitants came to characterise the god worshipped at the temple at Der el-Kala, they sometimes, it is clear, applied to
him grandiose appellations normally used for Iuppiter Optimus Maximus,
with his archaic temple on the Capitol in Rome. That in its turn must raise
a much wider question. For the same appellations were also used by them,
. Le Bas-Waddington, no.  CIG III  IGR III : κοίρανε κώμων.
. See, e.g., Le Bas-Waddington, no.  CIL III : Iovi Balmarcodi;  IGR III
: θεῷ Βαλμαρκῶδι; OGIS  IGR III ; Clermont-Ganneau, Rec. Arch. Or. 
():  and –, nos. –.
. W. Henzen, ILS III, no. , reprinted along with CIL III .
. Le Bas-Waddington, no.  CIL III ; OGIS  IGR III ; ClermontGanneau (n. ), , no. .
The Roman Coloniae of the Near East

far more frequently, for the much better-known deity whose temple was to
be found at Heliopolis: Iuppiter Optimus Maximus Heliopolitanus. He also
is sometimes described in simpler terms, for instance in the inscription of
.. / put up at Puteoli by the cultores Iovis Heliopolitani Berytenses qui
Puteolis consistunt.52
The question of how we should approach the nature of the Iuppiter of
Heliopolis will be considered below. For the moment it is essential to stress
that the entire situation, and the relation of the Berytenses to that cult, cannot
be understood correctly unless we accept the validity of the case argued by
A. H. M. Jones and ultimately refined by J.-P. Rey-Coquais that, as Strabo
clearly states, the Bekaa valley, and with it Heliopolis, was from the beginning part of the territory of Berytus.53 The cult of Iuppiter Optimus Maximus Heliopolitanus will thus have evolved as one observed primarily by the
Latin-speaking (and Roman-thinking) coloni of Berytus; and it was in this
period that the extraordinary group of temples was built, while the cult of
Iuppiter of Heliopolis steadily achieved an empire-wide fame. It was only
after two centuries, partially no doubt in recognition of that fame, and of
the fact of an urban complex existing around the temples, and certainly as
a product of the local tensions engendered by the civil war of –, that
Septimius Severus made Heliopolis a separate colonia. The fact is stated with
perfect clarity by Ulpian: Est et Heliopolitana (colonia) quae a divo Severo per belli
civilis occasionem Italicae coloniae rem publicam accepit.54 The fact that when this
status was granted, the newly acquired title, Colonia Iulia Augusta Felix Heliopolis, reflected the settlement of veterans by Augustus two centuries earlier
cannot weigh against the clear and irreducible facts. Pliny the Elder, who lists
the coloniae of Berytus (Nat. Hist. , ), Ptolemais under Claudius (, ) and
Caesarea under Vespasian (, ), knows Heliopolis only as a topographical
item (, ). There is no coinage of Heliopolis until the third century, and
no epigraphic evidence unambiguously denoting Heliopolis as a colonia until
the same period.
The colonia of Heliopolis, as it was in the third century, will be discussed
later (text to nn. – below). It should be accepted beyond all doubt
that in the first two centuries Heliopolis was a place (however described),
and a rapidly evolving cult centre, in the territory of Berytus. Inscriptions
from Heliopolis of this period which refer to the colonia, or to its decuriones
. ILS .
. See A. H. M. Jones, Cities of the Eastern Roman Provinces 2 (), – and n. ; cf.
J.-P. Rey-Coquais, IGLS VI: Baalbek et Beqa (), –; JRS  (): .
. Dig. , , , .

Rome and the East
or magistrates, must therefore be taken as allusions to the city government
of Berytus.55 The clearest and best-dated instance of such an allusion is a
statue base for M. Licinius Pompenna Potitus Urbanus, sacerdos of I(uppiter)
O(ptimus) M(aximus) H(eliopolitanus), granted the equus publicus by Hadrian,
and holder of a series of local offices: decurio, pontifex, agonothetes, duumvir
quinquennalis, flamen munerarius.56 These offices, and the shows referred to,
should be thought of as being those of the colonia of Berytus. In what specific
terms we should describe the status of Heliopolis in this period is not clear.
Following Rey-Coquais, however, we may note the apparent presence of a
pagus Augustus, seemingly making a dedication to the Dea Syria Nihathe(na)
at Niha, a place located south-west of Heliopolis on the other side of the
valley.57 Heliopolis was perhaps also a pagus. But whatever was the technical
term employed, we can in any case take it that, as Strabo makes clear, veterans were settled in the valley under Augustus. The fact that in the second
century some persons, including soldiers, began to record their origo as Heliopolis 58 does not prove that the place was already a colonia; rather, its emerging separate identity was a precondition of the grant subsequently made by
Septimius Severus.
That emerging identity must have both contributed to and been enhanced
by the extraordinary architectural development of the site and the widening
fame of the cult. Yet both the chronological sequence of the construction
and the sources of initiative (emperor or emperors? The colonia? The local
population? The cities of the wider Syrian region?) remain almost wholly
obscure.59 Even more profound problems, however, attend the question of
the origin and nature of the cult or cults of Heliopolis. The name of the place
has always suggested an association with Egypt, as already claimed by Lucian
(dea Syra ); and it may indeed derive from the domination of the region by
. E.g., IGLS VI , a dedication to Antoninus Pius by the colon[i s]plendidissimae
[col(oniae)]; –, probably of the second century; , King Agrippa (I or II), patrono
col.; , C. Iulius Sohaemus, patrono coloniae; also . Note also  from Niha: Q. Vesius
Patilianus, flamen au[g.], dec. Ber., quaestor col(onorum?) col(oniae?).
. IGLS VI .
. IGLS VI ; Rey-Coquais, JRS  (): . For Niha, see the important study by
Rey-Coquais, ‘‘Des montagnes au désert: Baetocécé, le pagus Augustus de Niha, la Ghouta à
l’est de Damas,’’ in E. Frézouls, ed., Sociétés urbaines, sociétés rurales dans l’Asie Mineure et la Syrie
hellénistiques et romaines (), –, emphasising the contrast with Heliopolis itself.
. IGLS VI,  and .
. IGLS VI –. For the architectural remains, see T. Wiegand, Baalbek I–III (–
); cf. M. Lyttelton, Baroque Architecture in Classical Antiquity (), chaps. –, and
F. Ragette, Baalbek ().
The Roman Coloniae of the Near East

the Ptolemies in the third century .. But there is only one inscription, and
no certain archaeological evidence, to show that there was a cult or construction there in the Hellenistic period.60 The temples there are of the imperial
period, from the first century .. to the third.
But what of the cult itself ? Can we confidently identify, behind the Romanised facade, a local ‘‘Semitic’’ cult, or even, as is now claimed, the cult
of a typical ‘‘Semitic’’ triad? That is the claim made in, and already incorporated in the title of, the invaluable and extraordinarily learned work on
this subject by Y. Hajjar:61 ‘‘Les dieux d’Héliopolis prennent dans les textes
épigraphiques et littéraires les noms de Jupiter, Vénus et Mercure ou leurs
équivalents grecs selon le cas. Mais il ne fait aucun doute que ces dénominations recouvrent des entités sémitiques avec Hadad, Atargatis et un parédre
mineur dont on ignore le nom indigène.’’
It will need a little time to demonstrate how frail is the logical basis of
this assertion, all too typical of what passes for the history of religion whenever the ‘‘Orient,’’ or the Near East, or anything supposedly ‘‘Semitic’’ comes
to be considered. If the assertion were merely that there was a cult on this
site in the Hellenistic period, that cannot be denied. For there is precisely
one fragmentary inscription of that period, including the words Διὸς ἱερῷ
and [εὐσ]έβειαν.62 We could then suppose that, as in the case of Der el-Kala,
where a cult of Theos Balmarcod was converted into one of Iuppiter Optimus
Maximus Balmarcod by Latin-speaking Roman veterans, so the cult of ‘‘Zeus’’
(nowhere given any Semitic appellation) at Heliopolis came in the same way
to be expressed as that of Iuppiter Optimus Maximus Heliopolitanus. But those
are the terms in which such a change must be described. For in speaking of
the identities of ancient deities, we are confronted by logical problems more
profound than simply the normal fact that what we can perceive is not these
divine entities themselves, but what our sources happen to say about them.
For the extra problem with the ancient pagan deities is that they were not
‘‘entities’’: they did not exist. What existed, and may be accessible to us, were
human beliefs and appellations, cult practices and temples. There is a perfectly valid sense in which ancient deities ‘‘were,’’ and can only have been,
what people at the time said they were.
If we follow strict logic, the ‘‘Semitic’’ triad of Heliopolis is the purest of
. F. Millar, ‘‘The Problem of Hellenistic Syria,’’ in A. Kuhrt and S. Sherwin-White,
eds., Hellenism and the East (),  ( chapter  of the present volume).
. Y. Hajjar, La triade d’Héliopolis-Baalbek I–II (); III (). The quotation is from
vol. II, –.
. Hajjar (n. ), I, no.  IGLS VI, .

Rome and the East
modern Semitising constructions. In the epigraphy of the Bekaa valley, not
a single allusion to Hadad occurs, still less an equation with Zeus or Iuppiter. Atargatis is indeed recorded once in the area; not, however, at Heliopolis
but in the hills on the opposite side of the valley, in the monumentum of Ochmaea at Niha, virgini vati Deae Syr(iae) Nihat(enae), or in Greek παρθένος θεᾶς
Ἀταργάτεις, put up by a veteranus named Sex. Allius Iullus (IGLS VI ).
The contrast with the epigraphy of Heliopolis itself which this inscription
presents is extremely important; for in this case a local deity is identified as
local, Nihatena, and as the ‘‘Syrian Goddess’’ and (in Greek) as Atargatis. No
such identifications are ever made by contemporary documents in relation
to the gods of Heliopolis. We are not entitled to assert that Iuppiter Optimus
Maximus Heliopolitanus was Hadad (or Hadaran); and all the less so because the
same Ochmaea (or Hocmaea) is also described at Niha, in another inscription, as virgo dei Hadaranis (IGLS VI ). At Niha Atargatis and Hadaran
may indeed have been worshipped as a divine couple.
There is in fact at Heliopolis no documentary equation of Venus with any
‘‘Semitic’’ goddess, or of Zeus or Mercurius with any ‘‘Semitic’’ god. But nor,
to make confusion worse, is there any real indication that these three deities
were conceived of as a triad at all (no ‘‘triad’’ of deities is represented on
the third-century coins of the city). There are in fact precisely two inscriptions from Heliopolis, both very fragmentary and heavily restored, where
we may read in the one case [I.O.M.,V.,] M., Diis Heliopol(itanis) (IGLS VI
), and in the other [I.O.] M., V. M., Diis Heliopol(itanis), where the first M
and V have not been seen since the early eighteenth century (IGLS VI ).
Otherwise I(uppiter) O(ptimus) M(aximus) H(eliopolitanus) is referred to alone
on nearly thirty inscriptions;63 or dedications are addressed in Greek to him
alone, as Διὶ Ἡλ[ι]οπολίτῃ or Διὶ μεγ[ίσ]τῳ Ἡλιοπολίτῃ (IGLS VI ;
cf. , also ). Venus/Aphrodite is much more rarely recorded (–,
). Mercurius/Hermes appears more frequently (IGLS VI –, ,
–; ?, ).
The inscriptional record from Heliopolis itself and the Bekaa valley thus
gives not the slightest justification for any notion of a triad; confirms the predominance of the localised Iuppiter Optimus Maximus who was worshipped,
mainly in Latin, by the colonial population; and also attests, if weakly, that
Venus and Mercurius, as Roman deities, were also worshipped there. From
what is said above, however, it will be clear that the whole context of the
evolution of the cults of Heliopolis is that of the territorium of the colonia of
. See IGLS VI, index v.
The Roman Coloniae of the Near East

Berytus. The reflection of the cult in Berytus itself is therefore of crucial
importance, in spite of the relatively small harvest of inscriptions from this
continuously inhabited city. The immensely learned tabulation and analysis
of the relevant inscriptions by Y. Hajjar do indeed offer valuable evidence,
and even at first sight offer some support for the notion of a triad. There are
in fact precisely two known inscriptions from Berytus where Iuppiter Optimus Maximus Heliopolitanus is named along with Venus and Mercurius, and
with these alone.64 A further inscription (mentioned above) offers only traces
of the M of [I.O.] M [H.], and follows that with Venus and Mercurius, but
also with Apollo, Deana [sic], Mars Sergit(ensis?), or Ceres, Proserpina, the
Fortuna of the Colonia, and the Fata.65 By contrast, eleven inscriptions from
Berytus and the surrounding area refer to I.O.M.H. alone,66 one to Venus
Domina, and two to Mercurius Dominus.67 An altar from the Museum of Beirut
might at first sight provide evidence for a syncretistic identification of Venus
Heliopolitana with the Syrian Goddess / Atargatis. But in fact, as Hajjar himself correctly argues, it shows rather a distinction between the two in the
mind of the worshipper. Two inscriptions appear on two sides of the altar:68
θεᾷ Ἀταργάτι / [σ]τ α (τίωνος) Γεράνων / Ἀρτεμίδι Φωσφόρω . . .
Veneri He/liopolitanae / et Deae Syriae / Geranensi / Deanae Luciferae
It seems clear that two deities are mentioned in both inscriptions, the Dea
Syria / Atargatis and Deana Lucifera / Artemis Phosphoros, while the Latin
also names a third, Venus Heliopolitana. Though the description of Berytus
as a ‘‘Latin island in a Semitic ocean’’ (text to n.  above) is far too simplistic, these inscriptions do indeed attest the complexities of the cultural and
religious context into which the colonists were settled. But precisely what
is distinctive about this area, marking it out from all other parts of the Roman Near East, is that among the other cultural heritages which were not
only still present, but active and evolving, a very strong Latin and Roman
element was firmly rooted, both on the coast and in the Bekaa valley. Whatever the nature or antiquity of the Zeus already worshipped at Heliopolis, he
now came to be characterised by appellations drawn from those of Iuppiter
Optimus Maximus in Rome, and came later to become known throughout the
. Hajjar (n. ), I, –, nos. –.
. Hajjar (n. ), I, , no. .
. Hajjar (n. ), I, –, nos. –, , .
. Hajjar (n. ), I, nos. –.
. Hajjar (n. ), I, no. .

Rome and the East
Empire as Iuppiter Optimus Maximus Heliopolitanus.69 Sometimes, though very
rarely indeed, the two other most characteristic deities of Romanised Heliopolis, Venus and Mercurius, were associated with him.70 But far more often
he appears alone. That wider diffusion raises other questions which cannot
be discussed here. In this context it is sufficient to emphasise the establishment in Berytus and its extensive territory of a Latin-speaking population
of Roman veterans, who worshipped Roman deities, even including a littleknown archaic goddess, Mater Matuta, under their Latin names. It was this
same context which enabled Valerius Probus to read in Berytus older Latin
works which had gone out of fashion in Rome (text to n.  above), and
which was to give rise to a long-surviving Roman cultural framework, of
which the law schools of Berytus were to be an important aspect. It is essential to stress that all our evidence up to the fourth century does indeed present
the teaching and learning of Roman law as an aspect of the ‘‘rather more Roman’’ city that Berytus was, and that we should speak of ‘‘law schools’’ rather
than of ‘‘the law school.’’ The second of these points is clearly and succinctly
expressed in the Expositio totius mundi et gentium of the middle of the fourth
century: Berytus civitas valde deliciosa et auditoria legum habens, per quam omnia
iudicia Romanorum ⟨stare videntur⟩.71
In the light of this it must be all the more of a puzzle that the most famous of all Roman jurists, Ulpian, was to come not from Berytus but from
the Phoenician-Greek city of Tyre, which had become a nominal colonia only
in his own life-time. It is time to turn to the later coloniae of the Roman Near
East, beginning with the distinctive group of three on the borders of, or at
the heart of, the Holy Land.
Ptolemais, Caesarea, Aelia Capitolina
These three coloniae, recently discussed very well by Benjamin Isaac,72 form a
clearly marked-out group, contrasting both with Berytus, on the one hand,
and with the much larger group of nominal coloniae of the Severan period,
and the following decades, on the other. It seems that all three foundations
. For the cult beyond its native areas, whose ramifications cannot be followed here,
see Hajjar (n. ), I, –, –.
. Hajjar (n. ), I, , no.  (Athens); , no.  (Zellhausen, Germania Superior),
a dedication by a praefectus cohortis from Berytus (the only examples).
. Expositio totius mundi et gentium  (Sources Chrétiennes , ed. J. Rougé [], ).
. B. Isaac, ‘‘Roman Colonies in Judaea: The Foundation of Aelia Capitolina,’’ Talanta
– (–):  (reprinted with revisions in B. Isaac, The Near East under Roman Rule,
Selected Papers []).
The Roman Coloniae of the Near East

were reflections of local conflicts, involving the Jews and their Roman rulers,
and indeed in the latter two cases there is no room for doubt on the matter.
In the case of Ptolemais, the earliest of them, a military context, and indeed
an actual veteran settlement, is probable, as Isaac has shown. But, perhaps
surprisingly, the actual process, or moment, of foundation cannot be located
within either of Josephus’ two detailed narratives of the period. Indeed, by
contrast with Berytus (n.  above), Josephus nowhere alludes to either Ptolemais or Caesarea as a colonia/ἀποικία. The fact that Ptolemais’ foundation as a
colonia (whatever that in practise meant) took place under Claudius (.. –
) is, however, attested by Pliny the Elder, writing in the s.73 As an event,
the foundation of the nova colonia is also reflected in a milestone of .. , on
the coast road from Antioch via Berytus to Ptolemais.74 Here too, of course,
whatever actual settlement of veterans took place did so in the context of
an ancient city, which in this instance had already acquired a Greek dynastic
name in the Hellenistic period.75 As ‘‘Ptolemais,’’ or, as it appears on coins of
Claudius, Γερμανικεῖς, it continued to mint Greek coins up to .. /, so
the colonia belongs in the last years of the reign. Thereafter the coins show
the city with its full Latin title Col(onia) Cla(udia) Stab(ilis) Ger(manica) Felix
Ptol(emais).76
There is also just sufficient evidence from the surrounding area to make
clear that veterans actually were settled. An inscription from beside the Roman road running north along the coast reads Imp. Ner. / Caesari / Col. Ptol. /
Veter. /, vici Nea Come et Gedru.77 The grammar is far from clear, but the refer. Plin., Nat. Hist. , : Colonia Claudi Caesaris Ptolemais, quae quondam Acce.
. P. Thomsen, ‘‘Die römischen Meilensteine der Provinzen Syria, Arabia und Palaestina,’’ ZDPV  (): , on , no. , a (from the Nahr el-Ghadir, south of Beirut): [Nero
Cl]audius [Caesar Au]g(ustus) Germanicus [trib. potest.] bis, cos. [designa]tus iterum [viam?] ab Antiochea [munivit? ad n]ovam colon[ia]m [Ptolemaid]a. . . . See R. G. Goodchild, ‘‘The Coast Road
of Phoenicia and Its Roman Milestones,’’ Berytus  (–):  and pl. xx.
. For the evidence, see Schürer, Vermes, and Millar, History II, –; cf. N. Kashtan,
‘‘Akko-Ptolemais: A Maritime Metropolis in Hellenistic and Early Roman Times,  ...–
 .., as Seen through the Literary Sources,’’ Med. Hist. Rev.  (): .
. See L. Kadman, The Coinage of Akko-Ptolemais (); see, for revisions of Kadman’s
views, H. Seyrig, ‘‘Le monnayage de Ptolmais en Phénicie,’’ RN  ():  Scripta Numismatica (), ; ‘‘Deux émissions coloniales incertaines,’’ RN  (): , who discusses,
on p. , the legend COL. C(onstans?) ST PTOL, found on some examples. I owe to Alla
Stein the information that the Greek coinage of Ptolemais terminates by .. /, rather
than / as Kadman states.
. M. Avi-Yonah, ‘‘Newly-Discovered Latin and Greek Inscriptions,’’ QDAP  ():
, , no.  AE , .

Rome and the East
ence to veterans is certain, and there is probably an allusion to two different
villages, one known by a Greek name and one by a Semitic one. There seems
no reason to dispute its identification with Talmudic GDRW / modern Kh.
Jidru, south-east of Akko. Also south-east of Akko we have a fragmentary
Latin inscription with the words pago and vicinal on separate lines 78—enough
to suggest the existence of a pagus of settlers here, as at Niha in the Bekaa
valley.
Some support, but far from wholly certain, for the idea of an actual settlement of veterans, is offered by the coinage of the city as a colonia; the earliest
coins, struck under Nero, name DIVOS CLAUD and have the normal symbol
of the founder with his plough marking out the boundary. Four legionary
standards are represented, but no agreement has been reached on the identification of the tiny numbers shown on them.79 It is none the less clear that
some actual settlement of veterans took place. But from the very slight evidence for the history of the city as a colonia nothing is really certain except
that it continued to mint coins with Latin legends identifying itself as a colonia into the reign of Valerian and Gallienus. There appear, however, to be no
Latin inscriptions from the city to illustrate public use of Latin there. The
real nature of the social and cultural mix produced by the process of foundation must remain mysterious. But here too we can be reasonably certain that
no wholesale removal or disturbance of the existing population took place;
the massacre of , Jews, and the imprisonment of many others, by the
Ptolemaeans in ..  must be seen as reflections of long-standing communal tensions, as they were in other cities of the region.80 On any construction
the foundation of a colonia here and at this moment remains a puzzle. It is
true, as Isaac points out, that there had been violent intercommunal clashes
between Jews and Samaritans in the procuratorship of Ventidius Cumanus
(probably .. –), which required the judicial intervention of the imperial legatus of Syria, Ummidius Quadratus, and led to the dismissal of the
procurator himself.81 But there is nothing to show that Quadratus, in visiting
. Avi-Yonah (n. ), , no.  AE , . J. Meyer, ‘‘A Centurial Stone from Shavei Tziyyon,’’ Scripta Classica Israelica  (–): , publishes what may be a centuriation
cippus from the territory of Akko. Note also the very full and suggestive study by Sh. Applebaum, ‘‘The Roman Colony of Ptolemais—‘Ake and Its Territory,’ ’’ in his Judaea in Hellenistic
and Roman Times: Historical and Archaeological Essays (), .
. Seyrig (n. ), –. Dr. C. J. Howgego kindly tells me that the numerals III, VI, X?,
and XI(?) can be read.
. Josephus, BJ , .
. Isaac (n. ), .
The Roman Coloniae of the Near East

Judaea to settle these disputes, had to bring any significant legionary forces
with him, or that coloni from Ptolemais subsequently played any role in repressing troubles in Judaea, or in the Jewish War itself (as a contingent from
Berytus had done earlier, after Herod’s death in  ..).82 There is also the
puzzle of the limited legal rights given to the new colonia. We would naturally
suppose that the Empire saw a progression from ‘‘real’’ coloniae, involving the
settlement of veterans and the granting of ius Italicum and full remission of
tribute, to ‘‘nominal’’ coloniae, which might receive full or partial rights depending on the indulgentia of the emperor. But in fact Ulpian is categorical as
to the absence of privileges on the part of this colonia (Dig. , , ): Ptolemaeensium enim colonia, quae inter Phoenicen et Palaestinam sita est, nihil praeter
nomen coloniae habet. Can it be that, far from being a fundamental conception
of Roman public law, integral to the overseas coloniae of the classical period,
the ius Italicum was a legal construct which evolved in the course of the Empire? 83 Or was it, as Miss Levick suggested to me, that Ptolemais had lost
some rights in the period between the foundation and the composition of
Ulpian’s De Censibus under Caracalla?
A closely comparable question presents itself in the case of the next colonia
founded in the Holy Land, Caesarea. Pliny the Elder (Nat. Hist. , ) carefully brings the history of the town right up to date: Stratonis Turris, eadem
Caesarea ab Herode rege condita, nunc Colonia Prima a Vespasiano Imperatore deducta. In doing so he reproduces almost in full the formal titulature of the
new colonia, as it was to appear on the coins, from the reign of Domitian onwards: Colonia Prima Flavia Augusta Caesarea.84 In this case there can be no
reason to doubt that the new status reflected the role of Caesarea as a military
base in the Jewish War, conducted of course by Vespasian himself up to the
moment of his coup d’état; indeed if we follow Josephus, it had been there
that Vespasian had first been hailed as emperor by his troops, in the summer
of .. .85
No source suggests that any deductio of veterans took place, and no legionary standards appear on the colonial coinage. As Isaac has pointed out, the
. Josephus, BJ , ; Ant. , .
. For this view, see F. T. Hinrichs, Die Geschichte der gromatischen Institutionen (),
–. Compare, however, J. Bleicken, ‘‘In provinciali solo dominium populi Romani est vel Caesaris. Zur Kolonisationspolitik der ausgehenden Republik und frühen Kaiserzeit,’’ Chiron 
(): , attributing the concept of ius Italicum to Augustus.
. See L. Kadman, The Coins of Caesarea Maritima ().
. BJ , –; see Isaac (n. ), –.

Rome and the East
language in which the jurist Paulus speaks of the status granted to the inhabitants by Vespasian clearly suggests a grant to an existing population (Dig. ,
, , ): Divus Vespasianus Caesarienses colonos fecit, non adiecto, ut et iuris Italici
essent, sed tributum his remisit capitis; sed divus Titus etiam solum immune factum
interpretatus est.
Ulpian confirms that the Colonia Caesariensis did not enjoy the ius Italicum (Dig. , , , ). It seems from the way that the jurists (though no earlier
source) express themselves that they saw coloniae as belonging potentially to
four possible grades: () no privileges; () partial remission of tributum; () full
remission of tributum; and () ius Italicum. Paulus’ account indicates that Vespasian, on granting the status of colonia, gave along with it only remission of
tributum capitis. It can be taken as certain that Titus’ interpretation, giving the
remission a wider definition, will have been in response to a submission by
the city, either transmitted by the legatus of Judaea or (more likely) brought
before him by an embassy.
By the standards of a middle-rank provincial town, the life of Caesarea is
illustrated, or potentially illustrated, by quite an extensive range of literature,
above all Christian. For it was there for several decades that Origen taught,
there that Pamphilus established his library, and there above all that Eusebius
lived, wrote, and served as bishop. Yet Eusebius for instance never refers to
Caesarea’s status as colonia. Nor, more significantly, does Gregorius of Neocaesarea, for whom, as we saw, Berytus was a ‘‘more Roman’’ city (text to n. 
above), but Caesarea was merely the seat of the governor of Syria Palaestina,
and by divine providence the place where he came to be a pupil of Origen.
Nothing in any of these sources would suggest to us that Caesarea was not
still a Greek city like any other. The same is largely true of inscriptions: witness the second-century inscription of a boxer from Aphrodisias cited earlier
(text to n.  above), in which the town appears as Caesarea Stratōnos, a curious
mixture of the two names which it had successively borne before becoming
a colonia. However, there is some slight inscriptional evidence reflecting the
new status of the place. One item is a remarkable Greek inscription of the
end of the second century from Mount Carmel, which also illustrates the attraction of the Zeus, or Iuppiter, of Heliopolis: Διὶ Ἡλιοπολείτῃ Καρμήλῳ /
Γ. Ἰούλ. Εὐτυχᾶς / κόλ(ων) Καισαρεύς. It is inscribed on the base of a statue
of which only the toes of one foot remain.86 The localisation of a cult of
the deity of Heliopolis on Mount Carmel raises complex problems of reli. M. Avi-Yonah, ‘‘Mount Carmel and the God of Baalbek,’’ IEJ  ():  AE
,  SEG XIV  Hajjar (n. ), I, no. , and pl. lxxxvi.
The Roman Coloniae of the Near East

gious history. For the present what is relevant is the self-identification of the
dedicator (evidently the descendant of a Greek family which had received
the citizenship at least some decades before the foundation of the colonia)
as a κόλ(ων) Καισαρεύς. The pseudo-Greek term is paralleled elsewhere in
the Near East: one example, from Palmyra, has already been noted (text to
n.  above).
The use of Latin in the colonia is now attested by a considerable number of inscriptions. There is a dedication by a lady named [Cleo?]patra,87
and more important, the well-known inscription of a man who, or whose
family, gained the citizenship in the Flavian period: M. Fl. Agrippam, pontif., IIviral. Col. I Fl. Aug. Caesareae, oratorem, ex dec. dec., pec. publ. (ILS 
Lehmann and Holum [n.  below], no. ). A single column from near
the city contains both a dedication to a legatus of the second or third century, by a duo[vir] or duo[viri], perhaps ex dec. dec. pec. publ.88 More significant
still is the pair of inscriptions of the first half of the third century found in
the theatre. One names Aur. Fl. Theophilus, an eques Romanus and decurio of
the metropolis, as being responsible for carrying out the decree of the decurions of the colony to honour Val. Calpurnianus, praefectus of Mesopotamia
and Osrhoene, pat(rono) metr(opolis - eos?) ex d(ecreto) d(ecurionum).89 The other
is a dedication to Aelius Iulianus, proc. Aug. n., also patr(ono) metr(opolis?) ex
d(ecreto) d(ecurionum).90 The inscriptions represent valuable epigraphic confirmation of a phenomenon which is better attested on third-century coinage,
namely the combination of the title colonia with the Greek status term for
major provincial cities, metropolis.91 If there was at any time a precise defini. AE ,  M. L. Lehmann and K. G. Holum, The Greek and Latin Inscriptions
of Caesarea Maritima (), no. . [Coloniae] Primae Fl. Aug. [Caesareae] / . . [Cleo]patra
mater eius hoc f(ieri) i(ussit); but see the better reading by Werner Eck cited there: [?Ex testamento—IIviral(is) (?) col.] Primae Fl(aviae) Aug(ustae) / [Caesareae—/—?Cleo]patra mater eius
her(es) [ facere iussit?].
. AE , a Lehmann and Holum (n. ), no. .
. AE , a W. Eck, ZPE  (): ff. Lehmann and Holum (n. ),
no. .
. AE , b Lehmann and Holum (n. ), no. . See now H. M. Cotton and
W. Eck, ‘‘A New Inscription from Caesarea Maritima and the Local Elite of Caesarea Maritima,’’ in L. Rutgers, ed., What Athens Has to Do with Jerusalem. Festschrift for Gideon Foerster
(), –, which discusses the use of Latin by members of the ruling families of Caesarea.
. For a useful sketch of the evidence, see A. Kindler, ‘‘The Status of Cities in the SyroPalestinian Area as Reflected in Their Coins,’’ Israel Numismatic Journal – (–): .

Rome and the East
tion of the conditions under which a city could call itself a metropolis, it is not
clear in our sources. It is at least certain that the title could be formally asked
for from the emperor and conferred by him.92 But it is equally clear that it
proliferated in the course of the empire and was not, for instance, confined
to one city in each province.
More important in the present context is the fact that it is a Greek status
title, which could be borne by coloniae as by purely Greek cities. In the case
of the two inscriptions just cited indeed, while the use of Latin has survived,
the actual title colonia has been displaced by metropolis. By contrast the coinage of Caesarea from Severus Alexander to Volusianus gives the title Col. I
F. Aug. F(elix?) C(?) Caes. Metrop.93 All the evidence combines to suggest that
the normal language of Caesarea was Greek. It was there that, according to
the Talmud of Jerusalem, Rabbi Levi bar Haitah, heard the Shema (‘‘Hear,
O Israel’’) being recited in Greek.94 He will have been in little danger of encountering the same phenomenon in Latin.
The last of the coloniae of this period was Aelia Capitolina, Jerusalem. Alone
of all the coloniae of the Near East, it arose from the destruction of an existing city and the wholesale replacement of its population by a new one. It
is not necessary here to enter into the question of the situation of Jerusalem after the Jewish War of .. , the problem of whether the foundation
of a colonia had been decided on before the Bar Kochba war of .. –
, or whether Jerusalem was ever captured by the rebels, and then recaptured by the Romans.95 Whatever the immediate background, there is no
reason to doubt Cassius Dio’s report that a new city called Aelia Capitolina
was founded in the place of the one which had been destroyed, and that the
temple was replaced by a pagan temple to ‘‘Zeus’’ ( Juppiter Capitolinus);96 a
gentile pagan population also took the place of a Jewish one;97 hence the first
bishop of non-Jewish origin was in office there soon after the re-foundation
(Eusebius, HE , , ). We can assume that in this case large-scale re-building
accompanied the foundation, and accounts of this process are offered later by
. See, e.g., F. Millar, The Emperor in the Roman World (), . No detailed study of
the use of the term is known to me.
. Kadman (n. ), –.
. Jerusalem Talmud, Sotah , , a.
. For the basic evidence, see Schürer, Vermes, and Millar, History I, –.
. Dio , , –. If he meant by saying καὶ εἰς τὸν τοῦ ναοῦ τοῦ θεοῦ τόπον ναὸν
τῷ Διὶ ἕτερον ἀντεγείραντος that the temple of Juppiter literally stood on the site of the
Temple, he was wrong.
. Zonaras , C; Malalas, Chron. .
The Roman Coloniae of the Near East

the Chronicon Paschale and by Joannes Malalas.98 Although much archaeological effort has been devoted in recent years to Roman and Byzantine Aelia,
the results remain both controversial and incompletely published, and there
is little that can be said with certainty as to the second-century city or the
quarters of the Legion X Fretensis which may have occupied part of it.99
There is nothing in our evidence to make clear whether the citizens of
the new colonia were drawn from veterans or from civilians; if the latter, a
new population must have been recruited, for contemporary sources make
clear that Jews were excluded.100 A colonia formed not from an existing city
but from new civilian settlers would however be unique in this period, and it
may be better to accept the colonial coins with vexilla of the legion X Fretensis as evidence that veterans were settled. If so, however, Aelia then provides
a unique combination of a legionary camp with a colonia of veterans from
the same legion.101 In either case, Aelia Capitolina and Mursa in Pannonia
Inferior represent the only two ‘‘genuine’’ coloniae of Hadrian’s reign, and the
last such coloniae in the history of Rome.102
With the new foundation there was associated an extensive programme
of road building in Judaea, and milestones from the relevant roads reflect,
though in Greek, the new status and name of Jerusalem: for instance ἀπὸ
κολ. Αἰλίας Καπιτωλ. μιλ. ε´.103 We would naturally suppose that the impact
of the foundation on the surrounding territory was considerable. Given the
model of Berytus, we might also suppose that Aelia, as (apparently) a new
veteran colonia, backed by a legionary camp, on a site whose original population had either been slaughtered or driven out, would have represented
another long-standing ‘‘Latin island.’’ If it did, however, there is remarkably
little evidence to show it. Christian sources of course have precisely focused
interests, but they do reveal that, apart from the temple of Juppiter, there
was also one of Venus, subsequently identified as the site of the Holy Sepulchre.104 But as an active expression of the city’s colonial status and Latin
character we have almost nothing to go on except its coins, where it ap. Chron. Pasch. , ; Malalas, Chron. . Note the translation by E. Jeffreys, M. Jeffreys, and R. Scott ().
. For a general account, see N. Avigad, Discovering Jerusalem (), –. See also
H. Geva, ‘‘The Camp of the Tenth Legion at Jerusalem: An Archaeological Reconstruction,’’
IEJ  (): ; R. Reich, ‘‘Four Notes on Jerusalem,’’ IEJ  (): , on –.
. Justin,  Apol. , ; Eusebius, HE , , , quoting Ariston of Pella.
. For these points, see Isaac (n. ), –.
. See M. Zahrnt, ‘‘Vermeintliche Kolonien des Kaisers Hadrian,’’ ZPE  (): .
. CIL III,   OGIS ; cf. IGR III  and Thomsen (n. ), –.
. E. D. Hunt, Holy Land Pilgrimage in the Later Roman Empire, .. – (), .

Rome and the East
pears as Col. Ael. Cap. and then, from Septimius Severus onwards, as Col. Ael.
Cap. Comm(odiana) Pia Felix.105 This latter name appears on just one inscription from the city, from near the southern wall of the Temple mount. Dating
from the reign of Severus, it provides the title [Colonia Ael]ia Kap(itolina)
Commo[diana].106
We are therefore left with a puzzle, on which the available evidence sheds
no real light. We have very little to illustrate the social composition, public
cults, or nature of local self-government (which we may assume was conducted by duumviri and a council of decuriones). Some relevant Latin inscriptions are indeed available: for instance a statue base in honour of Antoninus
Pius d(ecreto) d(ecurionum) (CIL III ); and another from the lintel of the
Roman gate underlying the Damascus Gate, naming the Colonia Aelia Capitolina.107 In what seems to be the only scholarly publication of it, the inscription is reported to read Co. Ael. Cap. d(ecreto) d(ecurionum).108
The availability of such fragments of evidence will hardly serve to obscure
our total ignorance of the dominant languages and the self-identification of
the inhabitants of what had been Jerusalem (and was frequently still called
Jerusalem by Christian writers) in its period as the Colonia Aelia Capitolina.
It is only when we reach the visit by the western pilgrim ‘‘Egeria’’ in the
s, that we learn that Christian services there were conducted in Greek,
with a regular translation into Aramaic, and a secondary translation, when
necessary, into Latin.109 Much had no doubt changed since Hadrian’s trium. See L. Kadman, The Coins of Aelia Capitolina ().
. AE ,  AE , ; cf. H. M. Cotton and W. Eck, ‘‘Ein Ehrenbogen für
Septimius Severus und seine Familie in Jerusalem,’’ in E. Dabrowa, ed., Donum Amicitiae.
Studies in Ancient History Offered by Friends and Colleagues on the Occasion of the Anniversary of
the Foundation of Ancient History in the Jagiellonian University of Kraków , –.
. Alluded to in various semi-popular reports, e.g., B. Mazar, The Mountain of the Lord
(),  (ascribed to Hadrian), , . A different impression, both of the location and
of the reading of the inscription, is given by N. Avigad, Discovering Jerusalem (), .
. R. W. Hamilton, ‘‘Excavations against the N. Wall, Jerusalem,’’ QDAP  ():
on –. For yet another impressionistic report of the area, see M. Magen, ‘‘Recovering
Roman Jerusalem: The Excavations beneath the Damascus Gate,’’ Bibl. Arch. Rev.  ():
. For a survey of the known Greek and Latin inscriptions, of all periods, from Jerusalem, see P. Thomsen, Die griechischen und lateinischen Inschriften der Stadt Jerusalem (), with
a Supplement in ZDPV  (): . For some updates, see B. Issac, ‘‘Inscriptions from
Jerusalem after the First Revolt’’ and ‘‘Epigraphic Remains from the Byzantine Period,’’ in
Y. Tsafrir and S. Safrai, eds., The History of Jerusalem: The Roman and Byzantine Periods (–
CE) (), –, – (in Hebrew).
. Peregrinatio Egeriae , –; see P. Maraval, ‘‘Égérie, journal de voyage’’ (Sources Chrétiennes , ).
The Roman Coloniae of the Near East

phant creation of a Roman colonia on the site, with a Roman constitution and
Roman gods. But the nature and evolution of that change escapes us entirely.
The Reign of Septimius Severus
With the ascent to the throne of Septimius Severus in ..  a quite new
phase opens in the nature of Roman coloniae in the Near East. As we will see,
it is not beyond possibility that his origins in Tripolitania, a Latin-speaking
area whose Punic sub-culture was a distant product of Phoenician colonisation, had some effect on this change. Equally, it may be relevant that his wife,
Iulia Domna, came from the city of Emesa in Syria, a place which contemporaries tended to characterise as ‘‘Phoenician,’’ even though it was situated
not on the coast but on the middle Orontes. Once again, the most difficult
problems of ethnic identity present themselves.
Two more obvious influences on the scale and nature of the formation of
new coloniae in this period were the civil war fought between Severus and
Pescennius Niger, the legatus of Syria, and the extension of the Roman frontier from the Euphrates to the Tigris. The title of colonia was to be conferred
on a whole series of places, mostly very little known or understood, in the
new provinces of Mesopotamia and Osrhoene. One place which received it
subsequently is, however, rather better known, namely Edessa/Orhai, which
until / was the capital of the small kingdom of Abgar.
The most clearly attested factor in this new phase was the use of the rank
or title of colonia as part of the repertoire of imperial rewards (with corresponding punishments), meted out to cities in respect of the attitude they
had taken in the civil war of –.110 Once again rabbinic literature happens to offer a remarkably vivid and apposite reflection of just this situation:
the people of a city are recorded as asking a king (emperor) to make their
city a colonia (QLWNY’), because they think it a good moment, as two of his
enemies have just been defeated.111
Contemporary sources make the connection between civil war and grants
of colonial status quite explicit, and in one case a combination of evidence
provides a precise date. Herodian records that there were local hostilities between Laodicea and Antioch and between Berytus and Tyre, and that Lao. For these processes, see R. Ziegler, ‘‘Antiochia, Laodicea und Sidon in der Politik
der Severer,’’ Chiron  (): .
. Sifre Deut. , , in the edition by L. Finkelstein (; repr. ). I owed the
reference to M. Goodman, State and Society in Roman Galilee, .. – (), , who
duly observes (in his n. ) the relevance of this to the civil war period.

Rome and the East
dicea and Tyre were quick to align themselves with Severus on the news of
his victory over Pescennius Niger in Cappadocia in ; Laodicea was then
sacked by Niger’s troops.112 This sack, and reports of the rewards which then
followed for Laodicea, appear in the sixth-century Chronicle of Malalas:
Severus granted senatorial rank to the leading men of the city, gave large
sums of money for public expenditures, built various public buildings, and
decreed that the city should have the status of metropolis and be named after
himself.113
The latter point is perhaps a confusion with Tyre, which was also rewarded
(see below). More significant is the fact that the rank of metropolis retained
its relevance in Malalas’ time; but the status of colonia, which Laodicea also
received, has dropped out of memory. But it is that which Ulpian, as a contemporary and a Roman jurist, had recorded, and related emphatically to the
civil war: Est et Laodicena colonia in Syria Coele, cui divus Severus ius Italicum ob
belli civilis merita concessit.114 Yet Malalas’ perspective was not entirely misleading, for a weight from Laodicea published by H. Seyrig shows that the year
/ counted as the th of the kolonia, but the th of the metropoliteia (the
status of metropolis).115 Severus had thus granted the title of metropolis on his
arrival in , but that of colonia only in , when he was again in the Syrian
region.
Both titles are reflected in the designation of an athlete from Laodicea,
in an inscription of .. , as κόλων Λα⟨ο⟩[δι]κεὺς μητροπολείτης κα[ὶ]
ἄλλων πόλεων πολείτης.116 The pseudo-Greek term κόλων is used again;
and the Romano-Greek public character of the city is reflected equally in its
coins. Under Caracalla one example has LAU COL METRO IUL;117 from then
on a variety of Latin terms appear, including AETERNUM BENEFICIUM
and ROMAE FEL. But a substantial series with the Latin/Greek legend COL
LAOD METROPOLEOS does not appear until the reign of Philip, ceasing
. Herodian , , –. See Ziegler (n. ), –.
. Malalas –.
. Dig. , , , . Cf. , , ,  (Paulus).
. H. Seyrig, ‘‘Un poids de Laodicée,’’ Syria  (): : ἔτους θ´ τῆς κολωνίας,
γι´ τῆς μητροπολιτείας, τ[οῦ] καὶ γνσ´. The latter figure, , establishes the date /.
Cf. the weight listed by H. Seyrig, Bull. Mus. Beyrouth  (): , no. : ἔτους δι´ τῆς
κολωνείας (hence .. /).
. IGR III  IGLS IV  Moretti, Iscrizioni agonistiche greche, no. .
. SNG, Danish National Museum: Syria, no. . It is curious that some of these colonial coins in Latin, as well as earlier Greek issues of the city, were counter-marked COL.
See C. J. Howgego, Greek Imperial Countermarks (), no. .
The Roman Coloniae of the Near East

in that of Gallus.118 With that, as it seems, all reflection of the colonial status
of the city also ceases. What is significant however is the clear dating of the
title metropolis to  and of colonia to , and the connection of both with
the posture adopted by the city in the civil war.
Apart from Antioch and Laodicea, the other pair of mutually hostile
neighbours which Herodian mentioned (text to n.  above) was Tyre,
which took the Severan side, and Berytus. Tyre too became a colonia, as we
will see below. But the reference also provides a perfectly clear and intelligible framework for the elevation of Heliopolis to the rank of an independent city with the status of colonia. That is to say that the urban conglomeration ( pagus?) which already existed around the long-famous temple
of Juppiter Optimus Maximus Heliopolitanus in the Bekaa valley (above)
was now detached from the territory of the colonia of Berytus on the coast.
The deployment of political favour, and disfavour, in this way finds an exact
parallel under Constantine, when the Emperor granted the port of Gaza,
Maiouma, city status as a reward for the Christian faith of its inhabitants. It
was thus detached from Gaza, whose population remained predominantly
pagan.119
It may have been merely that Berytus incurred displeasure and was punished by the loss of territory. Alternatively the people of Heliopolis, like the
inhabitants of the vicus Patavissensium in Dacia, may have petitioned Severus,
as Ulpian records (Dig. , , , ): In Dacia . . . Patavissensium vicus, qui a
divo Severo ius coloniae impetravit. But at any rate it is Ulpian who makes quite
clear the relevance of the civil war in the case of Heliopolis (Dig. , , , ):
Heliopolitana (colonia), quae a divo Severo per belli civilis occasionem Italicae coloniae
rem publicam accepit. Almost at the same moment as Ulpian was writing we
find Caracalla in ..  issuing a rescript about a case in which the res publica
Heliopolitanorum had come into possession of some property. It is interesting
(see below) that a party to the suit is called Sossianus (CJ , , ).
As we have seen, the area was already ‘‘colonial,’’ being part of the distinctive zone of Latin and Roman culture created by the settlement of legionaries
at Berytus and in its wide territorium. There is nothing in the least surprising or confusing in the fact that the formal title of the new colonia reflected
its historical origins in the veteran settlement under Augustus: Colonia Iulia
Augusta Felix Heliopolis. Not having been an independent city, the place had
of course minted no coins until Severus’ reign. From Severus to Gallienus we
. BMC Syria, –.
. Eusebius, VC , –; Sozomenus, HE , , –.

Rome and the East
duly find coins with COL (IUL AUG FEL) HEL.120 Under Philip and his son
the colonial coins also portray the standards of the two legions, the V Macedonica and VIII Augusta, which had taken part in the original settlement;
a few also incorporate the conventional scene of the founder ploughing.121
The coins, like the title of the colonia, should be read as conscious historical
reminiscence, or retrospection.
It is in keeping with the already established Latin, or Roman, character
of the area that we have a much better series of Latin inscriptions, from the
reign of Caracalla to the Tetrarchy, illustrating its colonial status than from
any other place. What is perhaps the earliest is probably a dedication to Caracalla ex dec(reto) dec(urionum) splendissimae col(oniae) Iul(iae) Aug(ustae) Fel(icis)
Hel(iopolis) (IGLS VI ). Another is securely dated to ..  (IGLS VI
); under Gordian a decurio of the colonia erected a ‘‘torch-bearing’’ statue
to mark his decurionate: statuam luciferam decurionatus sui hic conlo[ca]tam (IGLS
VI ); a dedication (IGLS VI ) and a milestone (IGLS VI )
date to the Tetrarchic period, as do newly published milestones from the
Heliopolis-Berytus road—Col[(onia) Iul(ia) Aug(usta) Fel(ix)] Hel(iopolis)—and
from the north-eastern limits of the city’s territory, near the sources of the
Orontes.122 There the series of local Latin inscriptions of the colonia ends,
along with the general decline of the epigraphic habit. There is also an undated dedication (IGLS VI ) by Baebius Aurelianus Dius dec. col. Hel. But
one native of the third-century colonia, L. Trebonius Sossianus (for the name,
text following n. ), who entered the Roman army and rose to the rank
of primus pilus, was to leave a record of himself and his city both in Rome
and at Philippopolis in Arabia. On the Janiculum in Rome he put up under
Gordian an inscription which perfectly expressed both the local and the ‘‘Roman’’ patriotism of Heliopolis: I.O.M.H., conservatori imperii d(omini) n(ostri)
Gordiani Pii Fel. Invicti Aug(usti), L. Trebonius Fab. Sossianus, colonia Heliopoli,
(centurio) frum(entarius) leg. III Fl(aviae) Gordianae p.p.123 A few years later the
same man, now a p(rimus) p(ilus) domo col. Hel., honoured C. Iulius Priscus,
. BMC Syria, –.
. L. Okamura, ‘‘Western Legions in Baalbek, Lebanon: Colonial Coins (.. –
) of the Philippi,’’ Historia  (): . I owe to Dr. C. J. Howgego the information
that the standards form part of a representation of what appears to be a group of cult statues,
two of which are holding the standards.
. Published by Ch. Ghadban, ‘‘Les frontières du territoire d’Héliopolis-Baalbek à
la lumière de nouveaux documents,’’ in La géographie administrative et politique d’Alexandre à
Mahomet (), , on  and  (Khirbet Choukan near the Orontes).
. ILS  Hajjar (n. ) I, no. , and II, pl. cxii.
The Roman Coloniae of the Near East

the brother of the emperor Philip, at their native city, Philippopolis (now
itself a colonia).124
Even in Heliopolis we would expect that both before and after the conferment of the new status Greek will have been current as well as Latin; and we
do duly find one Greek inscription which expresses the colonial character of
the place. A relief from the stairway of the main temple of Baalbek is accompanied by a brief inscription contained in a cartouche: Νάρκισος Κασίου,
βουλευτικὸς κολ(ωνείας) Ἡλ(ιουπόλεως).125 The inscription has sometimes
been taken to date to the first half of the second century, and hence to provide evidence that from the beginning Heliopolis was an independent colonia.
But the date is hypothetical, and a single brief inscription such as this cannot
possibly stand against the accumulation of other evidence. In fact it is to be
seen as a Greek reflection, of a familiar type, of the institutions of Severus’
new foundation.
The other place which Herodian (text to n.  above) records as having
displayed a timely loyalty to Severus in  was the city of Tyre. It can hardly
be doubted that its acquisition from Severus of the rank of colonia and the
privilege of ius Italicum came about precisely for that reason. But if that is so,
its proud citizen Domitius Ulpianus, certainly born some decades before the
change of status, chooses to record the fact in more grandiose language, reflecting the antiquity of the city and its long-standing relation with Rome: ut
est in Syria Phoenice splendidissima Tyriorum colonia, unde mihi origo est, nobilis regionibus, serie saeculorum antiquissima, armipotens, foederis quod cum Romanis percussit
tenacissima; huic enim divus Severus et imperator noster ob egregiam in rem publicam imperiumque Romanum insignem fidem ius Italicum dedit.126 From Ulpian’s
wording the date should not be earlier than .. , when Caracalla (imperator noster) became joint Augustus with Severus; and it very probably was
precisely , the same date as Laodicea.
Tyre, like other places, was now simultaneously a colonia and a metropolis.
The scanty epigraphy of the city, largely confined to grave inscriptions of the
late Roman period, offers very little confirmation of this, although one remarkable new discovery will be mentioned below. But the coins of Severus’
reign already show the legend SEP(timia) TURUS METROP(olis) COLONI(a),
and similar legends, with variations, continue up to the reign of Gallienus.
In the absence of all other evidence the fact that for a period under Elagabal
. ILS .
. IGLS VI .
. Dig. , , , pr; cf. , , ,  (Paulus): Eiusdem iuris et Tyriorum civitas a divis Severo
et Antonino facta est.

Rome and the East
(.. –) the city minted with the Latin legend TVRIORVM seems remarkably fragile evidence for the common notion that the status of colonia
was removed and then restored.127
The inscriptional evidence for the colonial status of Tyre is not extensive
but includes items of remarkable interest, complexity, and suggestiveness.
The earliest is certainly a Latin inscription from Lepcis Magna in Tripolitania,
the native city of Septimius Severus, itself a Phoenician/Punic city which
had been Romanised in the early Empire, and had become a colonia under
Trajan.128 Among the many inscriptions from there honouring Severus and
other members of his family, there is a statue base with the words [[P. Septimio
Getae nobilissimo Caesari]] Septimia Tyros Colonia Metropolis Phoenices et aliarum
civitatium.129 Severus’ younger son, Septimius Geta, held the title ‘‘Caesar’’
from ..  until he was made joint Augustus with his father and elder
brother in . His name was erased from inscriptions, including this one,
after his murder by his brother Caracalla, probably late in .. . The dedication of the original statue therefore belongs to the period .. –.
What is significant about the inscription is that it is a conscious reflection of the historic connections between the two cities which had been created by the Phoenician colonisation of the North African coast in the archaic
period, and was indeed reflected also in dedications by Lepcis, as the Colonia
Ulpia Traiana Aug(usta) Fidelis Lepcis Magna, put up in Tyre and honouring its
mother-city, before it in its turn became a colonia.130 Precisely for that reason
Septimia Tyros is not content here to be only Colonia and Metropolis Phoenices
but adds et aliarum civitatium. It is entirely consonant with this that among the
colonial coins of Tyre are ones portraying Dido supervising the foundation
of Carthage. Thus, in the dedication to Geta, Tyre, made a colonia by a citizen
of Lepcis, is recalling another and much older colonial relationship. The ancient glory of Tyre as a Phoenician city is also recalled on a colonial coin of
the reign of Gallienus, on which, apart from the normal COL TVRO METR
. So BMC Phoenicia, cxxv and –, followed, e.g., by H. Ingholt, Varia Tadmorea,
in Palmyre: bilan et perspectives (), , on , and K. Harl (n. ), . Alla Stein pointed out
to me that under Elagabal coins of Laodicea sometimes show ΛΑΟΔΙΚΕΩΝ or LAUDICEON, BMC Syria, –.
. For the background, see the very good account by A. R. Birley, The African Emperor,
Septimius Severus (), chaps. –.
. IRT .
. For what follows, see F. Millar, ‘‘The Phoenician Cities: A Case-Study of Hellenisation,’’ Proc. Camb. Philol. Soc.  (): , esp. – ( chapter  of the present volume).
For more dedications from Tyre, in Latin and Greek, see J.-P. Rey-Coquais, ‘‘Une double
dédicace de Lepcis Magna à Tyr,’’ in A. Mastino, ed., L’Africa Romana IV (), .
The Roman Coloniae of the Near East

there are the legends in Greek ΕΛΛΗ [ΝΕΣ ] and ΚΑΔΜΟΣ: the type recalls
the invention of the alphabet in Phoenicia, and its subsequent transmission
to Greece, by showing Kadmos giving a papyrus roll to three Greeks.
As was noted above (text following n. ), it could have been expected
that if a major Roman jurist originated from any of the coloniae of the Near
East, it would have been Berytus, or Heliopolis. But in fact the major jurist
of the classical period of Roman law, Domitius Ulpianus, came from Tyre.
His official career and his fame as a lawyer are duly reflected in a puzzling inscription engraved on a column found near the Roman arch of Tyre, which
itself belongs to the early third century.131 The lettering on the column may
be late Roman, and it is perhaps a re-inscription of a contemporary text. If it
is indeed a re-inscription, the fact that it is in Latin and concretely recalls the
colonial status of the third-century city will have made it a sort of historical
monument in itself. It is reported as reading:
Domitio Ulpiano, Praefecto
Praetorio, eminentissimo viro,
iurisconsulto, item Praefecto
Annonae Sacrae Urbis SEDERIA (SEBERIA?)
FELIX AUG [. . .] PRIOR COL METROPOL
P[A]TRIA
The city title at the end seems confused, and it could of course be that the
col(onia), metropol(is) named is not in fact Tyre but some other city. However,
Miss B. M. Levick, to whom I was very grateful for her comments on the
published text and photographs, suggested that instead of PRIOR in line 
we should read TYRIOR. SEDERIA in line  might be a personal name,
SEDEKIA(S) FELIX, described as AUG(ur?) or AUG(ustalis?) of the Tyrior(um)
col(onia) metropol(is). The last three terms would then be in the same order
as in the inscription from Lepcis Magna, though with the substitution of
a group genitive plural for the place-name. RIA in the last line might be
[MEMO]RIA. To go beyond that would be pure speculation. But Miss Levick
confirmed that the lettering seems to be of the fourth century.
More securely located in its historical context is a Greek inscription
of the mid-third century from Tyre, in honour of Septimius Odenathus:
. For the Roman arch of Tyre, see M. Chéhab, ‘‘Fouilles de Tyr. La nécropole I: l’arc
de triomphe,’’ Bull. Mus. Beyrouth  (); for the inscription, –, and pls. xiv–xv; cf.
AE , . See now M. Christol, ‘‘Entre la cité et l’empereur: Ulpien, Tyr et les empereurs de la dynastie sévérienne,’’ in F. Chausson and É. Wolff, eds., Consuetudinis Amor:
Fragments d’histoire romaine (II –VI e siècles) offerts à Jean-Pierre Callu (), .

Rome and the East
Σεπτίμ(ιον) Ὀδίναθον τὸν λαμπρότατ(ον), Σεπτιμία κολ(ωνία) Τύρος ἡ
μητρόπολις. The date should be the later s or earlier s, and the person concerned is the famous Odenathus of Palmyra, already a senator but
apparently not yet ὑπατικός.132 A Roman citizen and senator, Odenathus
came from a city which itself had been a Roman colonia for some decades (see
below). Tyre, however, in erecting this statue as an honour from the city, has
reverted to the use of Greek; and after the coinage ceases there is almost no
evidence either for the public use of Latin there or for its status as a colonia
being given emphasis, or even remembered. In the extensive writings of a
major third-century writer from Tyre, Porphyry, and in the biographical information about him, this status is nowhere alluded to: to Eunapius in the
later fourth century, he came from ‘‘Tyre, the foremost city of the ancient
Phoenicians’’ (Vit. Soph. ). The title metropolis, on the other hand, was still
in dispute between Tyre and Berytus in .. – (CJ , , ). It is clear
why it should have been that title which was in dispute, since metropolis implied a predominance over other cities which colonia did not. None the less,
there is an element of paradox in the appearance of this and not colonia in a
legal document of .. –. For exactly in those years we know that the
status of Tyre as a colonia could still be recorded: in the Acts of the Council of
Chalcedon, in a section relating to proceedings in , Tyre appears in this
guise: ἐν κολωνίᾳ Τύρῳ λαμπροτάτῃ μητροπόλει ὑπατικῇ.133 It had surely
been long since a mere title, which reflected nothing about the social and
cultural character of the city.134
By contrast, almost nothing is known of Septimius Severus’ transformation of Sebaste in Samaria into a colonia beyond Ulpian’s bare statement
(Dig. , , , ): Divus quoque Severus in Sebastenam civitatem coloniam deduxit. The latter expression is surely nominal, not implying that actual settlers
were brought there. Excavations there have provided evidence of signifi. For the inscription, see M. Chéhab, ‘‘Tyr a l’époque romaine,’’ Mél. Univ. St. Joseph
 (): , on –. For Septimius Odenathus, see M. Gawlikowski, ‘‘Les princes de
Palmyre,’’ Syria  (): .
. Acta Conc. Chalcedon. . Concilium Universale Chalcedonense I, , ed. E. Schwartz
(), .
. I would have argued that the fact that Eusebius describes his Greek text of the letter of Maximinus to Tyre as a translation from the Latin (HE , , –) is evidence for
the public use of Latin there in .. , but for the publication by S. Mitchell, ‘‘Maximinus and the Christians in .. : A New Latin Inscription,’’ JRS  (): , of a partial
text of the same proclamation, in Latin and addressed to the Colbassenses in the province
of Lycia and Pamphylia. The use of Latin is therefore not, in this context, distinctive of
colonial status.
The Roman Coloniae of the Near East

cant re-building—colonnades in the forum, a basilica, a columned street,
and a stadium—none precisely dated but possibly belonging to the Severan
period.135 The known inscriptions are extremely few; but one very fragmentary one seems to confirm that, as in other coloniae, there were duumviri, described by the normal Greek equivalent, στρατηγοί.136 The colonial coins at
any rate give some elements of the city’s title: COL(onia) L(ucia?) SEP(timia)
SEB(aste).137
That completes the list of the places within the already-established provincial area which received colonial status from Severus. What Sebaste did
to earn this award is not clear; but elsewhere the connection with timely
transfer to the Severan side in the civil war is explicitly or implicitly stated.
Rather different considerations inevitably apply to the remarkable list of coloniae which made their appearance in newly conquered Mesopotamia, one of
the least known and least understood areas of the ancient world.138
The initial creation of the new provinces of Mesopotamia and Osrhoene left
as an enclave the small kingdom of Abgar with its capital Edessa, which was
to become a colonia only under Caracalla (text to nn. – below). Here,
as will be seen, there is remarkable evidence. Elsewhere we know only bare
details. In the case of Nisibis at least, Cassius Dio indicates vaguely that a
status (ἀξίωμα) was given to the city at the moment of the formation of
the province; but elsewhere he elucidates this point by saying that Nisibis is
now ‘‘ours’’ and is regarded as a colony—νῦν μὲν ἡμετέρα ἐστὶ καὶ ἄποικος
ἡμῶν νομίζεται.139 The coinage confirms this, though the legends are without exception Greek: the title as given on the coins expands from ΚΟΛ
ΝΕΣΙΒΙ under Macrinus to ΙΟΥ(λία) ΣΕΠ(τιμία) ΚΟΛΩ(νία) ΝΕΣΙΒΙ
ΜΗΤ( ρόπολις) under Philip.140 Until recently, nothing more was known.
. J. W. Crowfoot, K. M. Kenyon, and E. L. Sukenik, The Buildings at Samaria (),
–.
. G. A. Reisner, C. G. Fisher, and D. G. Lyon, Harvard Excavations at Samaria, –
 (), , no. , and pl. c: [στρα]τηγούντων Ἀννίου Ῥο[ύφου . . .]. See Crowfoot
et al. (n. ), , no. .
. BMC Palestine, –, –.
. For this region, see L. Dilleman, Haute-Mésopotamie orientale et pays adjacents ();
N. Pigulevskaja, Les villes de l’état iranien aux époques parthe et sassanide (); D. Oates, Studies
in the Ancient History of Northern Iraq ().
. Dio , , ; , , . See Pigulevskaja (n. ), –, and for its later history as a
Roman colonia, C. S. Lightfoot, ‘‘Facts and Fictions: The Third Siege of Nisibis,’’ Historia 
(): .
. BMC Mesopotamia, cviii–ix, –.

Rome and the East
But now one of the fascinating documents which form a Greek-Syriac archive from mid-third-century Mesopotamia refers to a document drawn up
ἐν Σεπτιμίᾳ κολωνίᾳ μητροπόλι Νεσίβει, dating to .. .141 Ammianus’
narrative, which has so much to say of the fortunes of Nisibis in the fourth
century, gives no hint that it was a colonia. On the other cities of Mesopotamia
there is no literary evidence for this phase, and this status, at all. But the coins
show that Carrhae/Harran certainly, and Reshaina and Singara quite probably, became coloniae under Severus. Carrhae minted alternately in Greek and
Latin, the other two only in Greek.142 All three also claimed the title metropolis. In the case of Carrhae, one of the papyri (no. ) was drawn up in .. 
ἐν Αὐρηλίᾳ Κάρραις κολωνείᾳ μητροπόλει Μεσοποταμίας. The final publication of these papyri has shed some light on a potentially fascinating phase
in ancient cultural history. This evidence, fragmentary as it is, is still a reflection of the remarkable impact of Severus on the map of the Near East, or at
the very least on its toponomy.
The Third Century after Septimius Severus
Where Severus had led, his successors from the same dynasty, and their successors, followed. At least some aspects of the broader background to this
development are clear: firstly, with the acquisition of Mesopotamia the Near
East became one of the main areas of imperial military activity, an activity
stimulated in its turn by the Persian reaction of the s and after. Secondly,
the remaining emperors of the Severan dynasty, Caracalla (.. –), Elagabal (–), and Severus Alexander (–), descended on the maternal
side from a family from Emesa (which was to be another new colonia; see
below). Furthermore, Philip (.. –) came from a place in the northern Hauran which was also to become a colonia-cum-Greek city under the
name of Philippopolis.
The repeated presence of emperors in this region, and their deep involve. D. Feissel and J. Gascou, ‘‘Documents d’archives romains inédits du Moyen Euphrate
(IIIe siècle après J.-C.) III. Actes divers et letters (P.Euphr.  à ),’’ Journal des Savants :
–, no. .
. Carrhae/Harran: see G. F. Hill, JRS  (): –; BMC Mesopotamia, lxxxvii–
xciv; W. Cramer, ‘‘Harran,’’ RAC (), cols. –. (a) ΚΟΛΩΝΙΑΣ Μ Η ΚΑΡΩΝ
(Septimius Severus); (b) COL MET ANTONINIANA AUR and COL AUR METROPOLI
ANTONI (Caracalla); (c) ΜΗΤΡ ΚΟΛ ΚΑΡΡΗΝΩΝ (Gordian III). Reshaina: K.-O.
Castelin, ‘‘The Coinage of Rhesaena in Mesopotamia,’’ Num. Notes and Monog.  ():
ΣΕΠ ΚΟΛ ΡΗΣΑΙΝΗΣΙΩΝ (Decius). Singara: Hill, JRS  (): ; BMC Mesopotamia,
cxiii; –: ΑΥΡ ΣΕΠ ΚΟΛ ΣΙΝΓΑΡΑ.
The Roman Coloniae of the Near East

ment in military activity there, does serve in one way to explain, or at least
to provide a framework for, some of the grants of colonial status. A glance at
any map which specifies the formal status of cities in this region would show
a line of coloniae stretching along the desert frontier from Arabia to the Tigris:
Petra; Bostra; Philippopolis; Damascus; Palmyra, Dura-Europos(?); Carrhae;
Reshaina; Nisibis; Singara.143 But of course any notion that that status expressed any actual Roman military character, any role of these towns as fortresses in themselves, would be quite misleading. Furthermore, there were
other cities which also received this status, such as Sidon, Flavia Neapolis, or
Gaza, which can have been of no significance in external military terms; it
is just possible that the threat of disorder posed by the Samaritans was relevant in the case of Neapolis. By contrast, the title of colonia was not granted
to Seleucia in Pieria, which had served at least since the later first century
as the main Roman port for this region;144 explicit inscriptional evidence
from Cilicia shows supplies being delivered from there for the Roman armies
in Syria in the first half of the third century,145 surely through Seleucia. It
may seem equally surprising that the status of colonia, though granted, as we
would expect, to Antioch (see below), was not conferred on Apamea. For it
was not only a major city in its own right but was repeatedly the quarters
of the Legion II Parthica, on campaign with emperors far away from its base
at Albanum south of Rome.146 If military relevance had been the criterion,
we might also have expected that Zeugma on the Euphrates would also have
become a colonia.
We have to conclude that though these grants do indeed reflect the close
involvement of emperors with this region in the third century, they have
to be seen in each case as the expression of the vagaries of imperial favour.
Those made in the reign of Caracalla and after also have the further characteristic that they accompany, or follow, that Emperor’s grant of citizenship to
all the inhabitants of the Empire. With that, one fundamental distinction between coloniae and the Greek, or semi-Greek, cities which surrounded them
had been removed. In terms of citizenship there was now nothing to gain
from being a colonia. What other rights followed, and what exactly they sig. One modern map which does show this impact, though not completely, is map 
in N. G. L. Hammond, Atlas of the Greek and Roman World in Antiquity ().
. See the remarkable article by D. van Berchem, ‘‘Le port de Séleucie de Piérie et
l’infrastructure logistique des guerres parthiques,’’ Bonner Jahrb.  (): .
. AE , –. See H. Halfmann, Itinera Principum (), .
. See J.-C. Balty, Apamée (); ‘‘Nouvelles données sur l’armée romaine d’Orient
et les raides sassanides du milieu du IIIe siècle,’’ CRAI (): , and ‘‘Apamea in Syria in
the Second and Third Centuries ..,’’ JRS  (): .

Rome and the East
nified, depended on the terms of the grant in each case, which we do not
always know.
It may not be irrelevant that Ulpian from Tyre is the only jurist of this
period whom we know to have referred specifically to what moderns call
‘‘The Constitutio Antoniniana’’: in orbe Romano qui sunt ex constitutione imperatoris Antonini cives Romani effecti sunt. As the form of his allusion to him shows,
Ulpian was writing under Caracalla, in whose reign he composed, as it seems,
almost the whole of his vast corpus of work.147 Not much of the relevant
material incorporated in the Digest belongs more than a few years after that;
and, in fact, the examples of coloniae given by Ulpian and Paulus stop with
those of the reign of Caracalla. So, for instance, Ulpian can record the grant
of colonial status and ius Italicum to Emesa, by Caracalla, without (at least
in the quoted text) making clear that it was the home city of the Emperor’s
mother: Sed et Emisenae civitati Phoenices imperator noster ius coloniae dedit iurisque
Italici eam fecit.148 Until now nothing in the limited epigraphical material
from the site has reflected the new status of the city. But an inscription with
the names of Macrinus and Diadumenianus ends with RESTITUIT COL.
EMESENORUM. It comes from the south-western limits of the city territory, bordering on that of Heliopolis.149 The coins however are ‘‘colonial,’’ although they are all in Greek: we find ΕΜΙΣΩ ΚΟΛΩΝΙΑΣ or ΚΟΛΩΝ(ίας)
in – and , and ΜΗΤΡΟ ΚΟΛ (or ΜΗΤΡΟΚΟΛ?) ΕΜΙΣΩΝ under
Elagabal (.. –).150 These coins may thus represent the earliest attested
appearance of the hybrid Greek-Latin term μητροκολωνεία, attested a few
decades later also at Palmyra (see below). Whether it was intended in this case
to reflect Elagabal’s own derivation from the city is not clear. Beyond that,
our evidence fails us. The fourth-century literary evidence, which indicates
that Emesa was then in decline, gives no hint of its colonial status.151
Ulpian does not refer to Caracalla’s grant to Antioch. But Paulus, apparently writing after the Emperor’s death, does: Divus Antoninus Antiochenses
colonos fecit salvis tributis (Dig. , , , ). The grant was thus the purest of
. Dig. , , . See T. Honoré, Ulpian (), and the review article by F. Millar, ‘‘A
New Approach to the Roman Jurists,’’ JRS  (): – ( chapter  in F. Millar, Rome,
the Greek World, and the East II: Government, Society, and Culture in the Roman Empire).
. Dig. , , ,  (Ulpian); cf. , , ,  (Paulus): Imperator noster Antoninus civitatem
Emisenorum coloniam et iuris Italici fecit.
. See IGLS V: Émésène (). For the new inscription, see Ghadban (n. ), .
. BMC Syria, –. For the relevant legends under Uranius Antoninus in .. ,
see H. R. Baldus, Uranius Antoninus: Münzprägung und Geschichte (), –.
. See, e.g., Libanius, Ep. ; cf. H. Seyrig, ‘‘Caractères de l’histoire d’Émèse,’’ Syria
 (): .
The Roman Coloniae of the Near East

formalities, for salvis tributis will mean that no exemption from either tributum capitis or tributum soli was given. The coins, beginning under Elagabal, do however reflect the new status, showing legends, all in Greek, such
as ΑΝΤΙΟΧΕΩΝ ΜΗΤΡ ΚΟΛ, and ΑΝΤΙΟΧ ΜΗΤΡ ΚΟΛΩΝΙΑΣ.152 So
limited indeed was the impact of colonial status on the culture of Antioch
that Libanius, born in .. , could begin his Autobiography by refuting the
misconceived notion that his grandfather, presumably born in the middle
of the third century, had been an immigrant from Italy. The idea had arisen
because the grandfather had once composed a speech in Latin: ‘‘The fact is
that, although he was versed in Latin, he originated from nowhere else but
here.’’ 153 There seems to be nothing in the extensive fourth-century evidence relating to Antioch to suggest that the title was still remembered.154
However, another of the new documents from Mesopotamia (no. ) is a petition of .. , posted ἐν Ἀντιοχ(είᾳ) κολ(ωνίᾳ) μητροπόλει in the Hadrianic baths.155
Ulpian also records, immediately after Emesa, in his list of places with colonial status: Est et Palmyrena civitas in provincia Phoenice, prope barbaras gentes et
nationes collocata (Dig. , , , ). Ulpian’s description of the place is very
striking, for he must presumably be alluding to other unsettled peoples of the
Syrian desert west of the Euphrates. As we now know, Palmyrene settlement
extended a substantial distance west and north-west, and inscriptional evidence shows that their territory had common boundaries with Greek cities
to the west in the Orontes valley: Apamea (probably) and Emesa certainly.156
But the city also maintained at least outposts to the east along various routes
all the way to the Euphrates itself.157 If Ulpian were thinking of truly unsubdued Arab gentes and nationes, their territory must have lain to the north,
. BMC Syria, lviii–lxiii, –. See also G. Downey, ‘‘The Political Status of Roman
Antioch,’’ Berytus  (–): , and K. Butcher, ‘‘The Colonial Coinage of Antioch-onthe-Orontes c. .. –,’’ Num. Chron.  (): , curiously giving relatively little
attention to the colonial bronze coins as such, as opposed to imperial coins minted there.
. J. H. W. C. Liebeschuetz, Antioch: City and Imperial Administration in the Later Roman
Empire (), contains no allusion to colonial status.
. Libanius, Or. , , trans. A. F. Norman ().
. D. Feissel and J. Gascou, ‘‘Documents d’archives romains inédits du Moyen Euphrate
(IIIe siècle après J.-C.) I. Les petitions (P.Euphr.  à ),’’ Journal des Savants (): –.
. See D. Schlumberger, La Palmyrène du Nord-Ouest (), and H. Seyrig, ‘‘Bornes
frontières de la Palmyrène,’’ Syria  (): .
. M. Gawlikowski, ‘‘Palmyre et l’Euphrate,’’ Syria  (): . Note the excellent
map on .

Rome and the East
between Palmyra and Chalcis, or more probably to the south, where Palmyrene territory shaded off into the true desert of the Empty Quarter. At any
rate this is the aspect which Ulpian chooses to stress rather than any role in
the protection of long-distance trade with the East, or any function as a military strong point in relation to Parthia. By comparison with his allusions to
grants of colonial status in connection with the civil war of the s, he does
not in this case offer any explanation of how or by whom the title came to
be conferred. Nor does he make clear whether he would have thought of the
Palmyrena civitas itself as being somewhat barbara.
For us, however, though we remain entirely ignorant of the date or immediate context of the grant to Palmyra, the fact that Palmyra is the only
city in the entire Near East to offer a long series of public inscriptions in
both Greek and a Semitic language has the further consequence that it offers
the fullest documentary, and trilingual, reflection of the new status. Palmyra,
like Edessa (below), offers the image of a city which was ‘‘Semitic,’’ Greek,
and Roman all at once—or, at any rate, was so for the few decades before its
reconquest by Aurelian in .
Before the grant of colonial status, apparently under Severus or more
probably Caracalla, its formal character as a community had been merely
dual, Greek and Palmyrene. This character is best seen in the famous tax law
of the s, where the city is Ἁδριανὴ Πάλμυρα in Greek, HDRYN’ TDMR
(Tadmor) in Palmyrene. It has a βουλή and δῆμος / BWL’ WDMS; and officeholders with the Greek titles: proedros (given as the abstract noun proedria—
PLHDRWT’); archontes / ’RKWNY’; a grammateus / GRMṬWS. These terms
are all, as is obvious, merely transliterated into Palmyrene. But a further term,
dekaprōtoi, or perhaps better decuria (as Teixidor suggests), is not, being translated as ‘ŠRT’, or ‘‘group of ten.’’ 158
Palmyra, as it was in the last two-thirds of the second century, could thus
be seen as a place with the normal institutions of a Greek polis, but where
those institutions were, uniquely, also reflected in a long series of public inscriptions in a Semitic language. Like many a Greek city it had acquired also
a Roman imperial name, Hadriane; and the inscriptions also show a steady intrusion of Roman citizen names, borne by individuals. Then, from Caracalla
onwards, the standard ‘‘Roman’’ name borne by a Palmyrene is of a strange
. For the tax law, see CIS II, , . Note also J. Teixidor, ‘‘Le tarif de Palmyre I. Un
commentaire de la version palmyrénienne,’’ Aula Orientalis  (): , and J. F. Matthews,
‘‘The Tax Law of Palmyra: Evidence for Economic History in a City of the Roman East,’’
JRS  (): . For a discussion of the officials appearing in the tax law, see J. Teixidor,
Un port romain du désert: Palmyre (), –. Cf. also M. Zahrnt, ‘‘Zum Fiskalgesetz von
Palmyra und zur Geschichte der Stadt in hadrianischer Zeit,’’ ZPE  (): .
The Roman Coloniae of the Near East

form, with two ‘‘imperial’’ nomina, that is both ‘‘Iulius’’ and ‘‘Aurelius.’’ There
is no way of determining whether Caracalla’s universal grant of citizenship
or the specific grant of the status of colonia to Palmyra came first; but the latter
also seems to date to his reign.159
What is clear is that we have unambiguous evidence in the epigraphy of
Palmyra for the colonial status of the city and the adoption of a colonial constitution, headed by an annual pair of duumviri. But, apart from a couple of
milestones in Latin referring to the col. Pal., or col. Palm.,160 all the evidence is
in Greek or Palmyrene. The colonial status of the city is therefore expressed
in Palmyrene inscriptions either by transliteration (QLNY’) or by using the
normal Greek equivalent for the duumvir of a colonia, namely στρατηγός,
transliterated as ’SṬRṬG in Palmyrene. Once, however, we will see (below)
an attempt to produce in Greek a translation rather than equivalence.
Only a few of these items of evidence are dated to the decades immediately following the conferment of colonial status. The earliest is an honorific Greek inscription from the Great Colonnade, dating to .. /, and
recording Iulius Aurelius Seiba Ath[e]akabas and (Iulius Aurelius?) Titianus
Athenodorus as στρατηγούντων.161 Then there is a bilingual honorific inscription, also from a column in the Great Colonnade, dating to .. /,
and referring back to the service of Iulius Aurelius Zenobius, also called Zabdilas, on the occasion of the visit of the emperor Severus Alexander, in the
early s.162 In the Greek text he is referred to as στρατηγήσαντα, but in
the Palmyrene more fully as ’SṬRṬGLQLNY’. Earlier he had another office,
described as ἀγορανομήσαντα, surely in the context the equivalent of aedilis.
In this case the Palmyrene version gives a translation, RB ŠWQ, ‘‘controller
of the market’’ (souk). It is important to stress that Palmyrene was a living
language, whose use was not passively dependent on Graeco-Roman conceptions and terminology. This point is confirmed by a bilingual inscription (Inventaire X, ) from the Agora, in which Iulius Aurelius Malichus is
honoured as a past stratēgos and agoranomos of the colonia. The second office is
described in Palmyrene via an abstract noun, and the first probably is also:
. For this see D. Schlumberger, ‘‘Les gentilices romains des Palmyréniens,’’ Bull. d’Ét.
Or.  (–): , the latest detailed tabulation of the evidence; the article is also essential for the points which follow. Note also J. K. Stark, Personal Names in Palmyrene Inscriptions ().
. (a) CIS II, , : mil. xiv col. Pal.; (b) A. Poidebard, La trace de Rome dans le désert de Syrie (), : D. N. Imp. Caes. G. Val. Diocletiano p. . (?) Invicto Aug. Col. Palm.
mil. XXVIII.
. IGR III  J. Cantineau, Inventaire des Inscriptions de Palmyre III (), no. .
. IGR III  OGIS  CIS II.,  Inventaire III, no. .

Rome and the East
[B’SṬRṬGWT’ DY] QLNY’ WBRBNŠQWTH—‘‘in his stratēgia of the kolōneia and in his control of the market.’’
From October ..  we have a bilingual inscription honouring Iulius
Aurelius Ogga apparently as δυα[νδρικὸν, φιλοτείμ]ως στρατ[ηγήσαντα κα]ὶ
μαρτυρηθέν[τα . . .]. If the restorations are correct, the office of duumvir /
στρατηγός was referred to twice in different ways. Here again the Palmyrene version is not a simple transliteration or even a translation, but a slightly
different text: WŠPR LHWN B’SṬRṬGWTH—‘‘and gave them satisfaction
in his stratēgia.’’ 163 The remaining evidence, dating to the middle of the third
century, reflects the rise of Palmyra to independent power under the leadership of Septimius Odenathus. The spread to other leading Palmyrenes of
the Roman imperial nomen ‘‘Septimius’’ seems to be a result of this local
leadership, and not of an earlier set of grants by Septimius Severus.164 Thus
the person to whom most of the relevant texts refer, Septimius Worod,
seems to be the same as the ‘‘Aurelius Worod’’ attested in .. / as a Roman eques (which appears as ἱππικός / HPQ’) and city councillor of Palmyra
( βουλευτὴς Παλμυρηνός / BYLWṬ’ TDMRY’).165 With the assertion of independent power by Palmyra, this man rose to a series of important positions. In  a bilingual inscription from the Great Colonnade describes him
as [τὸν κράτιστ]ον ἐπίτροπ[ον Σεβαστοῦ δ]ουκηνάριον / QRṬYSṬS ’PTRP’.
At this point he is more probably still in imperial service. The man putting
up the inscription is a duumvir of the colonia, Iulius Aurelius Nebuzabadus,
[στρατ]ηγὸς [τῆς] λαμπροτάτης κολωνείας / ’STR⟨Ṭ⟩G’ DY QLNY’.166
Septimius Worod was also at some point aedilis and then duumvir of the
colonia, possibly before .. . For another bilingual inscription from the
Great Colonnade, whose Palmyrene text is almost entirely missing, and
whose precise date is uncertain, records that he had held these offices in
the past: λαμπρῶς στρατηγήσαντα καὶ ἀγορανομήσαντα τῆς αὐτῆς μητροκολωνείας. This new term μητροκολωνεία (perhaps first recorded at Emesa;
see text to n.  above) is a perfect expression of the hybrid character of
the more prominent cities in the Roman Near East in this period. The form
in which it is used here, however, is a function of the fact that the inscription has already mentioned Septimius Worod’s present post: [δι]κεοδότην
τῆς μητ[ ροκολω]νείας.167 This expression may or may not be intended as
. IGR III  CIS II.,  Inventaire III, no. .
. So Schlumberger (n. ), –.
. IGR III  OGIS  CIS II.,  Inventaire III, no. . For his identity,
see Schlumberger (n. ), .
. IGR III  CIS II.,  Inventaire III, no. .
. IGR III  CIS II.,  Inventaire III, no. .
The Roman Coloniae of the Near East

the equivalent of the Persian title argapet (governor), attributed to Septimius Worod on some other inscriptions of –, that is, the last couple
of years of the rule of Septimius Odenathus before his murder in .168
That raises problems which are not strictly relevant here. But it seems to be
from the same period that we have a fragmentary Greek dedication from
the monumental arch, using the expression ‘‘King of Kings,’’ and referring
to a victory over the Persians, put up by two men, one of whom seems to be
the same Septimius Worod with a more developed set of Roman Imperial
nomina: Ἰούλιος Αὐρήλιος [Σεπτί ]μιος Ο[ὐ]ο[ ρ]ώδης. The two dedicants are
described as ἀμφότεροι στρα[τηγοὶ τῆς λαμ]προτάτης [κ]ολω[ν]είας.169
It is not necessary here to pursue the question of the career of Septimius
Worod or the precise form of government and office-holding in Palmyra
during the brief period of its rise to power. Suffice to say that the city continued to designate itself as a κολωνεία / QLNY’ or μητροκολωνεία (no Palmyrene equivalent is yet attested), and to have stratēgoi among its annual officials up to the last few years of its fully independent existence. We might
have expected also that the coinage of Palmyra, as of the other coloniae of this
region, would contribute something to our understanding of its character.
But the small bronze coins of Palmyra, not certainly datable, reveal nothing
about its formal designation or status.170
The reconquest of the city by Aurelian in  took place at about the moment when city coinages generally ceased, and there is no reason to imagine
that minting there would have continued anyway. Inscriptions in Palmyrene
might have; but in fact it seems that they ended abruptly. The city was not
in fact destroyed; instead it continued, so far as we know, as a minor Greek
provincial place, with a Roman garrison.171 The few Greek inscriptions of
the fourth century and after, mainly Christian epitaphs, give no hint that
the city’s status as a colonia was still recalled. But there is at least one Latin
milestone of the Tetrarchic period to prove that colonial status survived the
reconquest (text to n.  above).
. IGR III  CIS II.,  Inventaire III, no. : ἐπίτροπον Σεβαστοῦ δουκηνάριον καὶ ἀργαπέτην / ’PTRP’ DQNR’ W’RGBṬ’. Cf. IGR III  CIS II.,  Inventaire III, no.  (.. ); and OGIS  IGR III  CIS II.,  Inventaire III,
no.  (.. ).
. IGR III  Inventaire III, no. . For restorations, see D. Schlumberger, ‘‘L’inscription d’Hérodien: remarques sur l’histoire des Princes de Palmyre,’’ Bull. d’Ét. Or.  (–
): , and M. Gawlikowski, ‘‘Les princes de Palmyre,’’ Syria  (): , on , no. .
. For a survey, see W. Szaivert, ‘‘Die Münzen von Palmyra,’’ in M. Ruprechtsberger,
ed., Palmyra, Geschichte, Kunst und Kultur der syrischen Oasenstadt (), .
. See F. Millar (n. ), .

Rome and the East
None the less, the evidence available shows that in the decades between
the grant of colonial status and the reconquest by Aurelian, the previously
‘‘Semitic,’’ Greek, or ‘‘Arab-Greek’’ city took on many ‘‘Roman’’ features. As
regards the widespread adoption of Roman names, largely in the anomalous
form ‘‘Iulius Aurelius . . . ,’’ we cannot clearly distinguish between the effects
of becoming a colonia, and the more general effects of the constitutio Antoniniana. But it is quite evident that the city adopted or received a new ‘‘colonial’’ constitution, with duumviri and aediles, under the Greek designation of
stratēgoi and agoranomoi.
Something similar can be traced in the case of the only place in the Near
East which has left a significant body of non-Jewish literary evidence in
a Semitic language, namely Edessa.172 When the new Roman provinces of
Mesopotamia and Osrhoene were formed in the s, the small kingdom of
King Abgar (IX?), with Edessa as its capital, remained as an enclave within
Osrhoene (AE , , .. , and , .. ). From Abgar’s reign
we have the one remaining substantial fragment of the sixth-century Syriac
Chronicle of Edessa, a vivid description of a flood in .. . This narrative
is far from revealing the entire structure of city self-government; but it does
show that the city, though dominated by a king, had its own communal officials, SPR’ D’WRHY (‘‘scribes of Orhai,’’ i.e., Edessa), and ‘‘superintendents
of the city’’ (ŠRYR’ DMDYNT’).173
Under Caracalla (.. –) there came a sudden transformation. King
Abgar IX (?)—or possibly his son, Abgar X (?)—was summoned by the Emperor and deposed, and the area became part of the Roman province of
Osrhoene. It seems clear that the exact date of this change was .. /.
For this is implied by the famous Syriac contract of sale, found at DuraEuropos, but written in Edessa in .. , and first published in .174 This
document is dated in a way which clearly points to a previous transforma. For a brief sketch, see F. Millar (n. ), –.
. Text and translation by R. Hallier, ‘‘Untersuchungen über die edessenische Chronik,’’ Texte und Untersuchungen IX (), – (trans.), – (text). The text is also reprinted in F. Rosenthal, ed., An Aramaic Handbook II. (), –. English translation by
J. Segal, Edessa: ‘‘The Blessed City’’ (), –.
. P. Dura ; revised text by J. Goldstein, ‘‘The Syriac Bill of Sale from DuraEuropos,’’ JNES  (): , text also in Rosenthal (n. ), – (Syriac only), and in
H. J. W. Drijvers, Old-Syriac (Edessean) Inscriptions (), –. Now reprinted, along with
the two new Syriac parchments of the same period from the Euphrates archive (see n. 
below), as P in H. J. W. Drijvers and J. F. Healey, The Old-Syriac Inscriptions of Edessa and
Osrhoene (), app. I.
The Roman Coloniae of the Near East

tion of the city’s status in  or : ‘‘in the month Iyár of the year  of the
former reckoning [the Seleucid era] and the year  . . . of the freedom. . . .’’
‘‘Freedom’’ (ḤRWRYH) here seems to refer both to the moment of provincialisation and to the apparently simultaneous elevation of Edessa to the
status of colonia. For the city is referred to as ‘‘the renowned Antonina [sic]
Edessa, Colonia Metropolis Aurelia Alexandria’’ (’NṬWNYN’ ’DS’ NṢYḤT’
QLWNY’ MTRPWLS ’WRLY’ ’LKSNDRY’). Colonia thus makes its appearance as a loan-word, slightly differently transliterated by comparison with
Aramaic and Palmyrene, in yet another Semitic language. As in various other
cases, it is associated with the purely Greek status appellation of metropolis
and with items of imperial nomenclature, of which the last will presumably have been acquired in the reign of Severus Alexander (.. –).
Here again, however, we are not presented with a mere change of name,
but with a new ‘‘colonial’’ constitution. For the document is also dated not
only ‘‘in the priesthood [BKMRWT’] of Marcus Aurelius Antiochus, Roman
knight [HPWS RHMWS—i.e., ἱππεὺς Ῥωμαῖος], son of Belšu,’’ but also by
the two stratēgoi—‘‘in the stratēgia [B’SṬRṬGWT’] of Marcus Aurelius Abgar,
Roman knight, son of Ma‘nu, grandson of Agga, and of Abgar, son of Hafsai,
grandson of Bar-Kmr.’’ On the verso one of the two stratēgoi/duumviri, Aurelius Abgar, appears as a witness, writing his own signature both in Syriac—
’WRLS ’BGR SṬRṬG’ ŠHD (witness)—and in Greek: Ἄβγαρος or Ἄβγαρ ὁ
στρ(ατηγός).175 That he could and did write both is a very significant fact.
The document shows throughout the impact of Roman citizenship and
Roman nomenclature. All the persons involved, both male and female, have
Roman names, with Semitic, or in one case Greek, cognomina. Almost the
same applies to the seller, Lucius(?) Aurelius Tiro, son of Bar B‘šmn, ḤRNY’
—‘‘Harranian,’’ that is, from Carrhae/Harran, also now a Roman colonia (text
to n.  above). But we also seem to find traces of the previous system
of city government of Edessa. For there is still a scribe (SPR’) in whose
hand the main body of the text is written, and who concludes it with the
words ‘‘I Marcus Aurelius Belšu, son of Moqimu, the scribe, have written this
document’’ (MRQWS ’WRLYWS BLŠW BR MQYMW SPR’ KTBT ŠTR’
HN’). Immediately above, another official, perhaps the supervisor of the archives, writes in Greek: Αὐρ. Μάννος ὁ ἐπὶ τοῦ ἱεροῦ καὶ τοῦ πολειτικοῦ
μ(α?)ρ(τυρῶ?). There follows a seal with the image of Gordian III (.. –
). The use of Greek, however, may serve to introduce one of the most
anomalous features in the history of Edessa as a Roman colonia. There appear to be no inscriptions reflecting this status. But, as elsewhere, there are
. Goldstein (n. ) denies that there is any visible trace of the final τ and ρ.

Rome and the East
coins, which, like those of most other Mesopotamian coloniae, are entirely in
Greek. The legends come in a variety of forms: ΚΟΛΩ(νία) Μ(ητρόπολις?)
Α(ὐ?)Ρ(ηλία?) or ΜΑΡ(κία?) ΕΔΕΣΣΑ; ΜΑΡ ΑΥ ΑΝΤ(ωνινιανὴ) ΚΟΛ
ΕΔΕΣΣΑ; ΜΗΤ ΚΟΛ ΕΔΕΣΣΑ, and so forth; the colonial coins begin
under Elagabal, or possibly Caracalla, and continue under Severus Alexander (.. –) and Gordian III (.. –).176 But then, still under Gordian III, we find coins with the Emperor on the obverse, and on the reverse
one figure seated on a podium and, before him, another standing figure,
wearing a high tiara and offering him a statuette. The identifications are
not left open to doubt; they are the Emperor himself and ‘‘King Abgar’’:
ΑΥΤΟΚ ΓΟΡΔΙΑΝΟΣ, ΑΒΓΑΡΟΣ ΒΑΣΙΛΕΥΣ. Other coins with Gordian on the obverse show King Abgar, riding a horse, on the reverse, or a bust
of him wearing a tiara.177 The coins thus offer a remarkable confirmation of
the view reached by von Gutschmid in  in studying the entries relating
to kings of Edessa which are scattered through the eighth-century Syriac
World-Chronicle, whose author should properly be identified as PseudoDionysius of Tell-Mahrē.178 The text of the chronicle is now available in a
modern edition, but one only usable by fluent readers of Syriac.179 In spite
of the excellence of von Gutschmid’s work, it should be stressed that the list
as established by him, and often subsequently quoted, was the result of extensive correction and adjustment; in particular, the chronicle plainly places
each of the later kings a couple of decades too early. The work needs to be
completely re-done in the light of subsequent documentary evidence. In the
mean time no reliance whatsoever can be placed on the numbering, or exact
dates or identities of the kings, as derived from this chronicle.
It is however clear that the list, however confused in the course of transmission, does embody reflections of a genuine succession of kings known
from contemporary evidence. Von Gutschmid’s attempt to make sense of it
was in itself entirely justified. He concluded, as regards the third century,
that in .. – there fell the thirty-five-year reign of a King Abgar, to
whom he gave the number IX, and whose full name was L. Aelius Septimius
Abgar. Without all the evidence being necessary here, this is essentially cor. BMC Mesopotamia, xciv–cvii, –.
. See H. Gesche, ‘‘Kaiser Gordian mit dem Pfeil in Edessa,’’ Jahrb. f. Num. u. Geldg.
 (): .
. A. von Gutschmid, ‘‘Untersuchungen über die Geschichte des Königreichs Osroene,’’ Mémoires de l’Académie Impériale des Sciences de Saint Pétersbourg, th ser., XXXV ().
. J.-B. Chabot, ed., ‘‘Incerti Auctoris Chronicon Pseudo-Dionysianum vulgo dictum,’’ Corpus Scriptorum Orientalium, Scriptores Syri III. I–II (–). For a very useful study
of this work, see W. Witakowski, The Syriac Chronicle of Pseudo-Dionysius of Tel-Mahrē ().
The Roman Coloniae of the Near East

rect; this is the ‘‘King Abgar’’ ruling when there was the flood of .. ,
recorded by the Chronicle of Edessa (text to n.  above); the ΒΑΣΙΛΕΥΣ
ΑΒΓΑΡΟΣ whom the coins show to have been a contemporary of Severus; and the ‘‘Septimius Abgar’’ whose regnum is attested in AE , , of
..  (text to n.  above).
Von Gutschmid concluded that there followed a joint reign of Abgar IX
and his son Severus Abgar (X), and then the nominal rule as ‘‘Titularkönig’’
of Ma‘nu son of Abgar, in .. –, followed finally by the restoration of
a genuine king, whom he thought was called Abgar (XI) Phrahates, son of
Ma‘nu.
The headings of the two extensive Syriac documents from the important new archive from Mesopotamia show both that there is a reflection of
a genuine sequence of events in all this, but also that it now requires radical readjustment.180 Firstly, the document designated P.Euphr.Syr. A is dated
..,181 in the Seleucid year , year  of Gordian and year  of ’LYWS
SPṬMYWS ’BGR MLK’ BR M‘NW PSGRYB’ BR ’BGR MLK’; that is ‘‘year 
of Aelius Septimius Abgar, son of Ma‘nu, crown prince, son of Abgar the
King.’’ As Teixidor suggested, this allows the reconstruction of a much simplified succession: firstly ‘‘the’’ Abgar, the contemporary of Severus (with the
possibility that it was he rather than a son called Abgar who was deposed
by Caracalla, as recorded by Dio , , a–2 ); then Abgar’s son Ma‘nu, left
with the status of ‘‘crown prince’’ (PSGRB’) for the twenty-six years /–
,182 then the restoration as king of Ma‘nu’s son, Aelius Septimius Abgar.
His second year coincided with Seleucid year , autumn –autumn ,
so his first coincided with autumn –autumn . It is thus he who appears
on coins along with Gordian III.
At that point the title colonia, which had been in regular use on the coins of
Edessa (text to n.  above) was dropped, for the name and title of the city
. J.Teixidor, ‘‘Les derniers rois d’Edesse d’après deux nouveaux documents syriaques,’’
ZPE  (): . I was extremely grateful to J. Teixidor and D. Feissel for a preliminary
text of the archive, and for permission to quote the ‘‘colonial’’ titles which appear there. For
my suggested reconstruction and dating of the entries in the Chronicle, see Millar (n. ),
app. .
. J. Teixidor, ‘‘Deux documents syriaques du IIIe siècle après J.-C., provenant du
Moyen Euphrate,’’ CRAI (): ; cf. S. Brock, ‘‘Some New Syriac Documents from
the Third Century ,’’ Aram .– (): ; B. Aggoula, ‘‘Studia Aramaica III,’’ Syria 
(): . This text is reprinted as P in Drijvers and J. F. Healey (n. ), app. I.
. It is worth noting that a Ma‘nu with the title PSGRYB’ is attested on a Syriac inscription from the citadel of Edessa; see H. J. W. Drijvers, Old-Syriac (Edessean) Inscriptions
(), no. . See Segal (n. ),  and pl. a.

Rome and the East
itself appears in the papyrus as ‘‘Orhai the fortified city which is the mother
of all the cities of Mesopotamia’’ (’RHY B’RS MDYNT’ DYT’ ’M’ DMDYNT’
KLHYN DBYT NHRYN).
The restoration thus certainly occurred a couple of years earlier than
had previously been supposed. It did not last long. As we saw, the Syriac
deed of sale from Dura-Europos shows that Edessa was again a colonia in
.. . But in fact colonial status had reappeared, and the king had disappeared, even earlier. For the Syriac document which Teixidor designated
B is dated .., year  of Gordian, Seleucid year , and year  ‘‘of the
liberation of the renowned Edessa Antonina Colonia Metropolis Aurelia
Alexandria’’ (DḤRWR’ D’NTWNYN’ ’DYS’ NṢYḤT’ QLWNY’ MTRPWLS
’WRLY’ ’LKSNDRY’)—therefore exactly as in the deed of sale of the following year.183
Inevitably therefore, the results of modern discussions of this tangled sequence of events now require radical revision.184 Some of the ‘‘colonial’’ coins
of Edessa under Gordian III may indeed date to after the restoration, and
coins of the reign of Decius (.. –) also have ΚΟΛ ΕΔΕΣΣΑ. Some
time not long after that coining will have ceased, along with that of all the
other cities of the Greek East.
The colonial constitution of the city continued, however. The Syriac
martyr act of Shmona and Guria, relating events which seem to have occurred in .. , gives a dating (para. ) by the stratēgia (B’SṬRṬYGWT’) of
Abba and Abgar; while the closely related martyr act of Habib the deacon,
describing comparable events which perhaps date to .. , gives a similar
local dating (para. ) as ‘‘in the stratēgia [B’SṬRṬGWT’] of Julius and Barak.’’
But this narrative also, like the record of the flood of .. , represents
‘‘superintendents’’ of the city (ŠRYR’ DMDYNT’, para. ), who play a role in
the persecution, reporting Habib’s activities to ‘‘the governor’’ (HGMWN’—
ἡγεμών, i.e., praeses).185 The standard Greek equivalent term for duumvir, stra. See Teixidor, ‘‘Un document syriaque de fermage de  ap. J.-C.,’’ Semitica –
(): –; cf. Brock and Aggoula (n. ). This document is reprinted as P in Drijvers
and J. F. Healey (n. ), app. I. Note the reconstruction of the dating mentioned in n. .
. See, e.g., R. Duval, Histoire d’Édesse (), –; A. R. Bellinger and C. B. Welles,
‘‘A Third-Century Contract of Sale from Edessa in Osrhoene,’’ YCS  (): –; E. Kirsten, ‘‘Edessa,’’ RAC IV (), –, on cols. –. For an excellent discussion of the
varied accounts (and names and dates of the kings) in the chronicles, and their relation to
the Syriac contract, see H. J. W. Drijvers, ‘‘Hatra, Palmyra and Edessa,’’ ANRW II. (),
–.
. For these texts, and translation, see F. C. Burkitt, Euphemia and the Goth (), –
.
The Roman Coloniae of the Near East

tēgos, had thus taken firm root in the political structure of the city. But there
does not seem to be any later Syriac literature relating to Edessa which shows
that same structure continuing.
That ends the list of the exotic coloniae of Mesopotamia; none, incidentally,
not even Nisibis, is referred to as a colonia by either Ulpian or Paulus. For
Ulpian, Palmyra, ‘‘situated near barbarous gentes and nationes,’’ was perhaps
exotic enough. His set of examples of coloniae (Dig. , , ) was written
under Caracalla (.. –); as for Paulus, the transmitted text of his enumeration (, , ) seems ambiguous as to whether Caracalla was still alive at
the moment of writing or not. At any rate neither makes any reference to the
colonial foundations of the reign of Elagabal (.. –), or later, a final
phase of which even less is known, and which may be summed up briefly
here. In very few cases is there any intelligible historical context for the grant
of colonial status. But since, of course, this status was already widespread
in the region, further grants may not require much explanation. Moreover,
with the s we are approaching even closer to the period when city minting and then the ‘‘epigraphic habit,’’ never so developed here as for instance
in North Africa or western Asia Minor, came wholly or largely to an end.
It seems that it was under Elagabal that Sidon became a colonia. Perhaps
surprisingly, its colonial coinage is consistently in Latin: AUR PIA SID COL
MET under Elagabal; COL AU(r) P(ia) METR SID, sometimes with AETERNU(m) BENEFI(cium), or COL AUR PIA MET SID under Severus Alexander (.. –).186 There are no inscriptions to attest the new status. But
by a fortunate accident there is a papyrus of .. , in which a pancratiast
records that he had been victorious in the ‘‘sacred iselastic oecumenical periporphyran isolympian contest’’ ἐν [κο]λωνίᾳ Σιδονίων πόλει.187 The nature
of the contest is entirely Greek; but none the less its Roman status is part of
the formal titulature of the city. It is also interesting to see this and other
evidence for the continuation of city athletic (and theatrical) festivals in the
period of the crisis of the Empire. But no further evidence is available to
illuminate the life of Sidon as a colonia.
It was also under Elagabal that the little-known city of Arca or Caesarea
ad Libanum, situated at the northern end of the Mount Lebanon chain, was
granted the rank of colonia. In this case the explanation seems simple, for it
. BMC Phoenicia, lxxxvii–cxvi, –. See N. Jidejian, Sidon through the Ages ().
Dr. Howgego suggested to me that both legends and style may be due to the influence of
Berytus.
. Sel. Pap. II, no. .

Rome and the East
was the native city of the Emperor’s uncle by marriage, Gessius Marcianus,
the husband of his aunt, Iulia Mammaea.188 Consequently the city was also
the native city of her son, Elagabal’s cousin and successor, Severus Alexander.189 But it was in fact early in Elagabal’s reign, soon after his coup d’état
in Syria in , that the elevation took place.190 The evidence in this case
is provided solely by the city coinage, which begins under Antoninus Pius
with the designation of the city as ΚΑΙΣΑΡΕΩΝ ΤΩΝ ΕΝ ΤΩ ΛΙΒΑΝΩ
or ΚΑΙΣΑΡΕΙΑΣ ΛΙΒΑΝΟΥ. Then in the Seleucid year , .. /,
coins appear with the Latin legend COL CESARIA LIBANI; under Severus
Alexander as Caesar, .. /, there is the variant COL CESA ITUR. Why
the historical connections of the city with the Itureans of Mount Lebanon
and Anti-Lebanon should have been recalled at this point is not known.
It seems similarly to have been under Elagabal that Petra received the rank
of colonia.191 Nothing whatsoever is known of the circumstances, but we find
on its coins from the reign of Elagabal the Latin legend COLONIA PETRA,
sometimes also written PETLA COLONI(A). Here again, therefore, the formerly Nabatean city, whose formal character had been for a century that of
a Greek city, with the title Ἁδριανὴ Πέτρα Μητρόπολις, was transformed
into a Roman colonia. By luck, given the very slight harvest of Greek and
Latin inscriptions from Petra, there is at least one which confirms the new
status. A statue base from the temenos of the Qasr Bint Far’un is inscribed in
Greek with the name of Valerius Iulianus, procurator of two emperors, honoured by the mētropolis and mētrokolōnia, and is dated by the two stratēgoi /
duumviri in office: Οὐαλέριον Ἰουλιανὸν τὸν / κράτιστον ἐπίτροπον τῶν /
Σεββ. ἡ μητρόπολις καὶ μητροκολ(ωνία) / τὸν ἑαυτῆς εὐεργέτην / διὰ [. . .]
ρ. Θεοδώρου καὶ Ἀριστείδου στρατηγῶν.192
The hybrid Latin-Greek word mētrokolōnia / μητροκολωνία thus appears
once again (cf. text to nn. – above). The inscription, modest as it is, is
. PIR 2, G ; see G. W. Bowersock, ‘‘Senators from the Near East,’’ in S. Panciera,
ed., Epigrafia e ordine senatorio II (), , on . I am not convinced, however, that Dig.
, , , shows that he had been adlected to the Senate, as against Dio , , , who says
merely that he had held procuratorships.
. So Aurelius Victor, Caes. , : Aurelio Alexandro, Syriae orto, cui duplex Caesarea et
Arce nomen est. Cf. HA, Sev. Alex. , ; , .
. BMC Phoenicia, lxxi–iii, –; see H. Seyrig, ‘‘Une monnaie de Césarée du Liban,’’
Syria  (): . Non vidi J. Starcky, Arca du Liban, Cahiers de l’Oronte  (–): .
. See S. Ben-Dor, ‘‘Petra Colonia,’’ Berytus  (): ; A. Spijkerman, Coins of the
Decapolis and Provincia Arabia (), –; G. W. Bowersock, Roman Arabia (), .
. Published by J. Starcky and C. M. Bennett in Ann. Dept. Ant. Jordan – (–):
, no. xii, and in Syria  (): , no. xiii. See now in Sartre’s IGLS XXI., no. .
The Roman Coloniae of the Near East

also sufficient to confirm that the normal restructuring of the city constitution took place here also.193 But in this as in almost all other respects the
subsequent history and nature of Petra as a Roman-Greek-Nabatean city in
the late Roman period remains a mystery. But note now the title appearing in
P. Petra  (..), lines –: ἐν Αὐγουστοκολωνίᾳ [Ἀ]ν τωνιανῇ ἐ[π]ι[σήμ [ῳ
καὶ] εὐ [. .]ε[ῖ μ]η [τρὶ] κ [ολωνιῶν] / Ἁδριανῇ Πέτρᾳ μητροπόλει τῆς Τρίτης
Παλαιστίνης Σαλουτ[αρίας].194
In the next reign, that of Severus Alexander (.. –), the other major
city of the province of Arabia, Bostra (known since Trajan’s acquisition of
Arabia as Νέα Τραιανὴ Βόστρα), followed suit, and gained the title colonia. Again, we have no historical context for the grant. Apart from a single
Greek inscription which records a dedication on behalf of Gordian III by
ἡ κολωνία, dated to the th year of the province (.. ),195 the sole
explicit evidence is provided by the city’s coins, which have the legends
COLONIA BOSTRA, N(ea) TR(aiana) A(lexandriana), COL BOSTRA and
COL METROPOLIS BOSTRA / BOSTRON / BOSTRENORUM.196 However the inscriptions of third- and even fourth-century Bostra do reflect
the existence of local offices with Latin titles. An undated Latin inscription (IGLS XIII ) mentions quaestoribus; but, more important, one of
/ is dated ἐπὶ κουαιστορείας (IGLS XIII ), while two men are described as ἀπὸ φλάμενος (IGLS XIII –). These texts are undated, but
surely derive from the third century. However there is nothing else in the
attested titulature of city officials in this period to reflect its colonial status,
not even the appearance of the term στρατηγός.197 We can be sure that the
normal language of Bostra, even for legal purposes, continued to be Greek:
the fact is explicitly attested in a papyrus from Oxyrhynchus (P. Oxy. )
recording the sale of a slave at ‘‘Bostra in Syria’’ in .. , κατὰ δίπλωμα
Ἑλληνικόν. Yet the document makes curiously insistent use of what seem to
. Starcky and Bennett (n. ), on pp.  and , refer to the two stratēgoi mentioned (as ’SRTGY’) in a Nabatean inscription from Hegra / Medain Saleh, and suggest
that these are in fact the annual stratēgoi / duumviri of Petra. The inscription is CIS II., ,
re-published by Jaussen and Savignac, Mission archéologique en Arabic I (), , no. .
Unfortunately the idea that this is a dating formula (‘‘in the days of ’’—‘L YMY) is removed
by the different reading in J. Cantineau. Le nabatéen II (), , no. viii.
. The Petra Papyri I, ed. J. Frösén, A. Arjava, and M. Lentinen ().
. IGLS XIII .
. See Spijkerman (n. ), –; A. Kindler, The Coinage of Bostra (); M. Sartre,
Bostra des origines à l’Islam (), .
. For the city’s institutions as reflected in the epigraphy of the region, see Sartre
(n. ), –.

Rome and the East
be Roman tribal names: φυλῆς Ἀν[. . .]—Aniensis? [. . .]εργια—Sergia? and
φυλῆς Ῥωμ . . ας—Romilia? 198 As elsewhere, we are confronted with complex mutual influences whose overall effect we cannot claim to understand.
Nothing suggests that further grants of colonial status followed in the reign
of Gordian III (.. –), in spite of the Persian campaign on which he
met his death. But several such grants seem to have come in the reign of his
successor Philip (.. –). In some instances the evidence is minimal. All
that we know of Damascus as a colonia is the legend COL DAMAS METR on
coins from Philip’s reign to that of Gallienus.199 Other cases are clearer. One
is Flavia Neapolis (present-day Nablus) in Samaria, which had been made
a city by Vespasian. Under Philip the new colonial status is reflected in a
remarkable variety of different Latin coin legends: COL SERG NEAPOL;
COL IUL NEAPOLI; NEAPOLIS COLON; NEAPOLI NEOCORO COL.
Then, under Gallus and Volusianus, there is COLON(IA) NEAPOLI(S).200 In
this case it so happens that a rabbinic source, a commentary on Lamentations, seems to provide a reflection of the new rank: ‘‘After Jerusalem was
destroyed Caesarea became a metropolis [MṬRWPWLYN], Antipatris a city
[MDYNH] and Neapolis a colonia [QLWNYY’].’’ 201 The Midrash on Lamentations is thought to date to the fifth century. If so, it is interesting, and consonant with the other evidence, that Caesarea’s status as a metropolis should
be more visible than that as colonia; but the more recent grant, of colonial
status to Neapolis still attracts attention.
. See H. M. Cotton, W. Cockle, and F. Millar, ‘‘The Papyrology of the Roman Near
East: A Survey,’’ JRS  (): nos. –. On P. Bostra , see now J. Gascou, ‘‘Unités administratives locales et fonctionnaires romains. Les données des nouveaux papyrus du Moyen
Euphrate et d’Arabie,’’ and H. M. Cotton, ‘‘Appendix: Administrative Divisions in Arabia,’’
in W. Eck, ed., Lokale Autonomie und römische Ordnungsmacht in den kaiserzeitlichen Provinzen
vom .–. Jh. (), – and –, respectively.
. BMC Syria, lxxiv–v, –.
. BMC Palestine, xxv–xxxiv, –. See K. W. Harl, ‘‘The Coinage of Neapolis in
Samaria,’’ Am. Num. Soc. Mus. Notes  (): .
. Midrash Ekha Rabbati . (ed. Buber. , ). Translated by A. Cohen in Midrash
Rahbah, Lamentations, ed. H. Freedman and M. Simon (), , where the references to
Antipatris and Neapolis in Buber’s text are rejected, without however any reason being
given. I was grateful to Aharon Oppenheimer for confirmation that Buber’s text is acceptable. Antipatris had been founded by Herod the Great—see Schürer, Vermes, and Millar,
History II (), —but a remarkable confirmation of the historicity of the context is
offered by coins of Antipatris from the reign of Caracalla or Elagabal, probably the latter,
with the legend Μ ΑΥΡ ΑΝΤ ΑΝΤΙΠΑΤΡΙΣ. See Y. Meshorer, The City Coins of EretzIsrael and the Decapolis (), no. .
The Roman Coloniae of the Near East

In parallel with that there is of course the much better-known case of
Philippopolis in the northern Hauran, founded as a city with the rank of
colonia by its most famous citizen, Philip ‘‘the Arab.’’ The place will perhaps
have been a substantial village before its elevation, though nothing is in fact
known of its earlier name (probably, but not certainly, related to its present
name, Shahba), of its character as an urban conglomeration, or of its degree
of self-government; it may even be (as Segal, n.  below, suggests) that the
new city was built on a virgin site. This lack of information is unfortunate but
also surprising, for the physical remains of the substantial villages, and occasional towns (Canatha and Soueida/Dionysias), of this region, with their rich
crop of inscriptions illustrating village self-government, allied with intensive study in recent decades, have made this one of the best-known regions
of the Roman Near East.202
The Emperor, M. Iulius Philippus, born in about .. , was the son
of a local notable of equestrian rank, Iulius Marinus, both of whose sons
embarked on prominent equestrian careers.203 The notion that we are concerned with a family of low-class brigands (Epit. de Caes. , ) can safely be
dismissed; more probably they were quite substantial landowners. The refoundation of the place as a city with the rank of colonia had a considerable
impact both locally, in the wider Syrian region, and in the historical tradition on Philip. In the town itself we find the building, in the very distinctive
black basalt of the region, of a shrine dedicated to the now deified father of
the Emperor, Marinus: θεὸς Μαρίνος or Divus Marinus.204 The theatre of Philippopolis, also in black basalt, and situated close to the sanctuary, seems to
have been built at the same time.205 So, probably, was the main street, along
with the walls and gates.206 The constitution of the new city is not illustrated
. See, of course, the well-known article by G. M. Harper, ‘‘Village Administration
in the Roman Province of Syria,’’ YCS  (): ; H. I. MacAdam, ‘‘Epigraphy and Village Life in Southern Syria in the Roman and Early Byzantine Periods,’’ Berytus  ():
, and Studies in the History of the Roman Province of Arabia: The Northern Sector (); and,
above all, J.-M. Dentzer, ed., Le Hauran I.– (–).
. PIR 2 I , , . For the family and the place, see now G. Amer and M. Gawlikowski, ‘‘Le sanctuaire impérial de Philippopolis,’’ Dam. Mitt.  (): .
. Le Bas-Waddington, nos. – IGR IV –; AE ,  (reference
only): given in full by Amer and Gawlikowski (n. ), , Divo Marino eqq. ala Celerum
Philippianae.
. See P. Coupel and E. Frézouls, Le théatre de Philippopolis en Arabie ().
. There has been no comprehensive analysis of these buildings since the accounts
by H. C. Butter, AAES II: Architecture and Other Arts (), –, and by R. E. Brünnow
and A. von Domaszewski, Die Provincia Arabia III (), –. For a survey, necessarily

Rome and the East
by any documentary evidence; but a Greek inscription contains a dedication
for the safety of Philip and his son by three bouleutai, προεδρίᾳ Μαρρίνου,
ἔτους πρώτου τῆς πόλεως.207 Coins of the reign of Philip (but not after)
have the Greek legend ΦΙΛΙΠΠΟΠΟΛΙΤΩΝ ΚΟΛΩΝΙΑΣ.208 It was very
likely of this new foundation that someone at Sakkaia (later Maximianopolis, now Shaqqa), a few miles away, was thinking when he put up a verse
epigram ending with the words εὐτυχίτω ἡ κολωνία—‘‘Good luck to the
colonia.’’ 209
Rather further afield, the author of the Thirteenth Sibylline Oracle, written a few years later, probably in Emesa, noticed the adornment of this city
and of one other new colonia in the vicinity, Bostra, but not specifically their
new formal status: ‘‘Now bedeck yourselves, cities of the Arabs, with temples
and stadiums, agoras and streets, glorious wealth, and statues of gold and
silver and ivory . . . Bostra and Philippopolis.’’ 210 The new foundation also
achieved a mention in Aurelius Victor’s brief account of the reign of Philip,
written over a century later: conditoque apud Arabiam Philippopoli oppido (Caes.
, ). Here too, as so often, and almost universally in the fourth century, the
specific status of colonia has dropped out of sight.
With that the list of coloniae in the Near East is almost complete: the very
uncertain cases of Ascalon, Gadara, and Gerasa, none of which produced any
colonial coins, need not be discussed in detail.211 There remains, firstly, Durabased on published material, see A. Segal, ‘‘Town Planning and Architecture in Provincia
Arabia: The Cities along the Via Traiana Nova in the st–rd Centuries ..,’’ BAR Int. Ser.
 (): –.
. Le Bas-Waddington, no.  IGR III .
. BMC Arabia, xli–ii, –; Spijkerman (n. ), –.
. IGR III . Note the inscription of ..  recording that someone had dedicated a statue of the Tyche of Sakkaia, τύχην Μεγάλην Σακκαίας τῇ κυρίᾳ πατρίδι (AE
,  bis).
. Orac. Sib. , –. See A. T. Olmstead, ‘‘The Mid-Third Century of the Christian
Era,’’ Class. Philol.  (): , on –; Baldus (n. ), –.
. (a) Ascalon. The evidence consists solely of a papyrus of .. , BGU, no.  L. Mitteis and U. Wilcken, Chrestomathie, no. , .: ἐν κολωνίᾳ Ἀσκ[άλωνι?] τῇ πιστῇ
καὶ ἐλευθέρᾳ. See Schürer, Vermes, and Millar, History I, –. (b) Gadara. The sole evidence is CIL III   (Byblos): col. Valen. Gadara. See ZDPV  (): –, and
Schürer, Vermes, and Millar, History II, –. (c) Gerasa. There are two inscriptions from
the city which refer, or may refer, to colonial status, but not certainly that of Gerasa itself.
C. H. Kraeling, Gerasa City of the Decapolis (), , no. : Colonia Aur. Antoniniana; ,
The Roman Coloniae of the Near East

Europos, which also never minted its own coins, but where a papyrus contract of divorce of .. , almost the latest document from the history of
the town, gives it a title which possibly reflects both its foundation in the
early Hellenistic period and a new Roman status: ἐν κολωνείᾳ Εὐρω
π [αίων
Σελεύκου] Νεικ ά τ ο ρο ς [τ]ῇ ἱερᾷ [καὶ] ἀ [σύλ]ῳ καὶ α [ὐ]τ ονόμ[ῳ]. The title
is enough to give substance to a couple of other references, one with the
loan-word κόλωνες, in the plural, among the inscriptions in Dura, suggesting that the title had been conferred here too, by which emperor we do
not know.212
The same three Greek titles, ‘‘sacred, inviolate, and autonomous,’’ are used
to describe the city of Gaza in an inscription from Portus near Rome, dating
to the reign of Gordian III.213 It seems clear that the place was not yet a colonia. But a lead weight from Gaza itself shows that it too did become one at
some point: κολωνίας Γάζης, ἐπὶ Ἡρῴδου Διοφάντου.214 It is perhaps paradoxical that this relatively obscure place should not only have gained such an
honour, but should also have been the location of quite extensive narratives
relating to the fourth and fifth centuries, which duly confirm that it then had
a colonial constitution. The first is the brilliantly evocative Vita of Hilarion
by Jerome, covering the first half of the fourth century.215 Being in Latin, it
can of course deploy the correct terminology. We thus find that the chief
official of the city is a Gazensis duumvir (ch. ). Confirmation is offered by
Sozomenus’ Ecclesiastical History, which speaks of στρατηγοί (, ); and perhaps by Mark the Deacon’s Life of the Bishop Porphyry, which speaks of ‘‘the
two leading men,’’ τῶν δύο πρωτευόντων, named Timotheos and Epiphanios, along with a dēmekdikos and some eirēnarchoi (), in office in the last
couple of years of the century. However, two chapters later () he speaks of
no. : στρατηγὸν κα[ὶ πά]τρωνα τῆς κ[ολω]νείας. See Kraeling, . Later excavations
do not seem to have produced any further inscriptions bearing on this topic.
. P. Dura , .–. Cf. F. Cumont, Fouilles de Dura-Europos (), , no. : πιάκλα
[εἰ]σὶ κολ(ώνων?); , no. : κόλωνες, βουλευταὶ καὶ εἱερεῖς; and Excavations at DuraEuropos III (), , no.  (statue base for Julia Domna): Αὐρηλ(ιανῶν) Ἀντωνινιανῶν
Εὐρωπαίων ἡ βουλή; V (), , no. : κολωνιοδ[ουρανός?].
. IGR I . See C. A. M. Glucker, ‘‘The City of Gaza in the Roman and Byzantine
Periods,’’ BAR Int. Ser.  (): .
. IGR III ; Glucker (n. ), , no. .
. Migne, PL XXXIII, cols. –. English translation in R. J. Deferrari, ed., Early
Christian Biographies (Fathers of the Church XV, ), –; text with introduction, Italian translation, and notes in C. Mohrmann, ed., Vite dei Santi IV (). See also Glucker
(n. ), ; H. Grégoire and M.-A. Kugener, eds., Marc le Diacre, Vie de Porphyre ().

Rome and the East
three prōteuontes; if this is not a simple error, it could mean the two duumviri
(?) and the ecdicus, or defensor.
Neither Jerome nor Mark the Deacon explicitly refers to the city as a
colonia; if their evidence suggests that the city still had a colonial constitution, it is only implicitly. So far as is known, the city had never minted any
colonial coins; and if any coloniae were still benefitting from remission of direct taxation, Mark’s narrative categorically implies () that Gaza was not
among them.216
Conclusion
It is not entirely unsuitable that this survey of the coloniae of the Roman Near
East should end on so uncertain a note. From one point of view the discussion might serve to emphasise how fragile, limited, and disparate is the
evidence for the physical nature, social composition, self-government, and
collective images of cities in this region under Roman rule. From another,
however, it might serve to suggest how considerable the changes brought
about by Roman rule were. Even if we ignore fundamental issues of social history such as the extension of cultivation, the growth of substantial
villages in many regions, and the efflorescence of places calling themselves
towns, Roman rule profoundly affected the personal and collective identities
by which people lived. Quite outside the list of places which became coloniae,
Latin place-names, from ‘‘Julias,’’ ‘‘Livias,’’ and ‘‘Tiberias’’ onwards, were scattered liberally across the map of the Near East, while Latin personal names,
and then the Roman tria nomina, found their way into common use, not only
in Greek but in Semitic languages.
From the strictly linguistic point of view it is also revealing that the
Near East was a zone in which Latin, Greek, and Semitic languages operated in a complex set of interrelationships. So colonia and colonus could become κολωνία, or κολωνεία, and κόλων, and these in their turn QLNY’, or
QLWNY’, and QWLWN. In the form of a loan-word colonia and its cognates found their way into several different Semitic languages: Hebrew, Jewish Aramaic, Palmyrene, and Syriac. Since Nabatean remained in use until
the fourth century, it is not impossible that the word might one day turn up
here too.
The main wave of grants of colonial status, beginning under Septimius
Severus, immediately preceded, or accompanied, Caracalla’s grant of Roman
. The historical status of the entire narrative is, however, a matter of controversy,
not least in this section. See also Glucker (n. ), , n. .
The Roman Coloniae of the Near East

citizenship for all the inhabitants of the Empire. The status thus automatically lost one aspect of its significance, to be absorbed into the wider range
of Romanising influences in the Greek East, a mutual interaction which was
to give birth to the ‘‘Roman’’ Empire ruled from Constantinople. Whether
taxation privileges were maintained is obscure, as is the real significance and
real applicability of ius Italicum, as a right conferring on the possession of
land in the provinces the status of full ownership in Roman law. Any discussion of this right ought at least to start from the fact that one of the two
fullest accounts of its distribution (both very brief ) comes from Domitius
Ulpianus, a native of a pseudo-colonia, Tyre, of the Severan period, and the
other from his contemporary Paulus. That in itself is a puzzle, for if there
is one item of social and cultural history which stands out from this survey
of the coloniae, it is that it was primarily the earliest of them, Berytus, with
Heliopolis as part of its territory, later turned into an independent colonia,
which created a real and enduring island of Romanitas and of the use of Latin
in the Near East. (But recently published inscriptions show that the same
was true of Caesarea, at any rate until the end of the third century.) It was
a function of that ‘‘Roman’’ character that Berytus gave birth to law schools
(not a law school) which continued to attract students until the Byzantine
period. It was the city as a whole which fostered the study of Roman law.
Thus Eunapius (VS ) describes Anatolius, a native of the place who rose
to the praetorian prefecture in the s, as ‘‘rising to the summit of what is
called legal learning, as having as his native city Berytus, which is regarded
as the mother of such studies.’’
Even Berytus, however, founded in and absorbing an existing GraecoPhoenician city, had from the beginning minted coins like a Greek city; in
continuing to do so it expressed its identity as a city in a way unlike all except one of the Roman coloniae of the Latin-speaking part of the Empire (see
n.  above). The evidence, however scattered and inadequate, shows that the
later coloniae, which nearly all minted coins, might do so either in Latin or in
Greek. It shows equally clearly that colonia, or κολωνία, competed as a status
designation with the specifically Greek term metropolis.
Thus to all appearances—and appearances are all that the evidence allows
us to grasp—the coloniae of this region, Berytus apart, were gradually reabsorbed into their environment, as Greek cities among others, with only
the structure of their magistracies, and occasional reappearances of the word
κολωνία in their formal titulature to single them out. For Jerome, recording Paula’s pilgrimage to the Holy Land in the s, of all the cities (Tyre,
Sidon, Caesarea, Aelia itself ) which were or had been coloniae, only Berytus
still earned the title Romana colonia (Ep. , ). Very little, so far as we know,

Rome and the East
remained of that brief but very distinctive phase, occupying roughly the first
half of the third century, which momentarily gave a very precise meaning to
the notion of ‘‘Graeco-Roman’’ civilisation as represented by cities, and also
saw both the largest number of individual city mints (under Septimius Severus for the Greek East as a whole, and under Elagabal for the Syro-Palestinian
region) and the greatest overall volume of production of city coinages. But
in the third quarter of the century all such minting ceased.217 Thereafter our
quite extensive evidence for the life of this region and its cities provides no
more than the most slight and erratic evidence to suggest that the status of
colonia was still remembered. Only Berytus, the earliest and most profoundly
Romanised of these colonial foundations, continued to play its historic role
as the source of Roman law for the early Byzantine world.
. See A. Johnston, ‘‘Greek Imperial Statistics: A Commentary,’’ RN  () and Harl
(n. ).
 
Latin in the Epigraphy
of the Roman Near East *
Introduction
In approaching the complex problems of the role of Latin in the epigraphy
of the Roman Near East, it seems appropriate to begin with by far the most
famous of all Roman inscriptions, the titulus on the Cross. It is, of course, not
typical of what we normally refer to as ‘‘inscriptions,’’ since it was written—
presumably painted—on a non-permanent material, wood, and was thereby
designed for the information of the public over only a short space of time, no
more than a few hours. But it did share with all inscriptions, whether temporary or permanent, the intended function of making information available
to whoever could read it, or have it read to them. It has moreover an exceptional importance for students of epigraphy, because it alone, of all the
millions of inscriptions put up in the Roman Empire, is the subject of four
separate reports in literary sources. It also serves as an object warning, as regards the effective transfer of information: for not one of the four reports
coincides exactly with any of the others; and only one of them records even
the elementary fact that the inscription was put up in three languages. The
texts of the relevant passages from the four Gospels are as follows:
καὶ ἦν ἡ ἐπιγραφὴ τῆς αἰτίας αὐτοῦ ἐπιγεγραμμένη, ὁ βασιλεὺς τῶν
Ἰουδαίων. (Mk. :)
καὶ ἐπέθηκαν ἐπάνω τῆς κεφαλῆς αὐτοῦ τὴν αἰτίαν αὐτοῦ γεγραμμένην, οὗτός ἐστιν Ἰησοῦς ὁ βασιλεὺς τῶν Ἰουδαίων. (Mt. :)
ἦν δὲ καὶ ἐπιγραφὴ ἐπ᾿ αὐτῷ, ὁ βασιλεὺς τῶν Ἰουδαίων οὗτος. (Lk.
:)
* First published in H. Solin, O. Salomies, and U.-M. Liertz, eds., Acta Colloquii Epigraphici
Latini. Helsinki – Sept.  (Commentationes Humanarum Litterarum , ), –.


Rome and the East
ἔγραψε δὲ καὶ τίτλον ὁ Πιλάτος, καὶ ἔθηκεν ἐπὶ τοῦ σταυροῦ. ἦν δὲ
γεγραμμένον, Ἰησοῦς ὁ Ναζωραῖος ὁ βασιλεὺς τῶν Ἰουδαίων. τοῦτον οὖν τὸν τίτλον πολλοὶ ἀνέγνωσαν τῶν Ἰουδαίων, ὅτι ἐγγὺς ἦν ὁ
τόπος τῆς πόλεως, ὅπου ἐσταυρώθη ὁ Ἰησοῦς. καὶ ἦν γεγραμμένον
Ἑβραϊστί, Ῥωμαϊστί, Ἑλληνιστί. ( Jn. :–)
It will be seen at a glance that by far the most detailed and informative of
the four accounts is that of John. That of itself will of course not prove that
this narrative should be preferred to the Synoptic versions, from which it
differs in almost every particular, including above all the relation of the Crucifixion to Passover. But in fact there are many reasons for believing that the
‘‘historical Jesus’’ is best reached through John’s Gospel.1
All four reports of what the inscription said are transmitted to us in Greek,
by authors whose origin (whether Jewish or gentile), current place of residence (whether in Judaea or elsewhere), or date of writing are all entirely
matters of speculation. But the fact that they did write in Greek is of course
a perfectly accurate reflection of the centrality of Greek as the primary language of communication throughout the Roman Near East. Given that role,
it also functioned as a medium of transmission between the various Semitic
languages used in the region on the one hand and Latin on the other. This role
is perfectly exemplified, for instance, in a perishable medium, in the census
return of ..  from the archive of Babatha, in which both Babatha’s subscription, originally in Aramaic, and that of the Roman praefectus, originally
in Latin, appear in Greek translation.2
In the three Synoptic Gospels, however, the coexistence of Latin, Greek,
and a Semitic language, as symbolised on the Cross, has dropped out of view,
to leave only three different reports in Greek of what the inscription said. It
is not only in that respect that John’s report is more complex and detailed.
It is also he alone who records, whether veridically or not, that Jesus’ origin
in Nazareth was indicated; and only he who reflects the currency of transliterated Latin official terms in everyday provincial Greek, in calling the inscription a τίτλος, or titulus: in this sense a notice written on any form of
material, for public display.
. See for the best discussion of the issues, E. Bickerman, ‘‘Utilitas Crucis,’’ RHR 
():  ( Studies in Jewish and Christian History III [], ); and for the significance of
John’s Gospel, F. Millar, ‘‘Reflections on the Trials of Jesus,’’ in P. R. Davies and R. T. White,
eds., A Tribute to Geza Vermes: Essays on Jewish and Christian Literature and History (), 
( chapter  of the present volume), and R. Lane Fox, The Unauthorised Version: Truth and
Fiction in the Bible (), ff.
. N. Lewis, The Documents from the Bar-Kokhba Period in the Cave of Letters: Greek Papyri
(), no. .
Latin Epigraphy

More significant, however, is John’s report that the titulus was written
in three languages, Ἑβραϊστί, Ῥωμαϊστί, Ἑλληνιστί. The latter two are of
course straightforward, namely Latin and Greek. But it remains systematically uncertain what is meant by Ἑβραϊστί.The problem arises because, while
it is certain that both Hebrew and Aramaic were in use among the Jewish
population of Judaea, there was no clear and distinctive terminology for distinguishing between them. Ἑβραϊστί or κατὰ τὴν τῶν Ἑβραίων διάλεκτον
can be used by Josephus and New Testament writers to describe words which
are in fact in Aramaic.3 As regards the various dialects of Aramaic used by
gentiles in the Near East, there is a fair degree of consistency among observers
writing in Greek or Latin. They speak of Suroi (or sometimes Assurioi) in
Greek and of Syri in Latin, and of people speaking suristi or ‘‘in the Syrian
language.’’ We would not expect from them a detailed knowledge of dialects,
or of regional variations. But it is significant for their global view of what
we call ‘‘Aramaic’’ that outsiders applied terms cognate with ‘‘Syrian’’ to local
dialects in use throughout the entire region; thus to Pliny the Elder, Hierapolis in northern Syria ‘‘is called Mabog by the Syri’’; Josephus records that
Palmyra was still called ‘‘Thadamora’’ among the Suroi; and Eusebius says that
Petra was called ‘‘Rekem’’ among the Assurioi.4
But although the Aramaic used by Jews in Judaea was also just another
dialect of the Aramaic which outsiders thought of as ‘‘the language of the
Syrians,’’ and was distinct from Hebrew, it does not seem that, within a Jewish context, there was a clear distinction of nomenclature between the two.
Indeed it is clear that there was not. So when either Paul in Jerusalem, or
Josephus outside the walls of the city during the siege of .. , is described
as making a speech to the people τῇ Ἑβραΐδι διαλέκτῳ, or ἑβραΐζων, we
cannot be sure whether Hebrew or Aramaic is meant. Either is in fact possible.5 But if either could be used as a spoken language, a fortiori either might
have been used, along with Latin and Greek, in the formal context of the
titulus on the Cross.
In the Near East, therefore, Latin entered into a more complex set of linguistic relationships than we can trace in any other area of the Empire. By and
. For the most detailed survey, see still Schürer, Vermes, and Millar, History II, –.
For the confusion, see esp. , n. . See also H. M. Cotton’s new edition of the famous lines
of P.Yadin , lines –: ἐγράφη δ[ὲ] Ἑληνεστὶ διὰ τ[ὸ] [ἡ]μᾶς μὴ εὐρηκ [έ ]ναι Ἑβραεστὶ
ἐ[γγρ]ᾴψασθαι, in Y. Yadin, J. C. Greenfield, A. Yardeni, and B. Levine, The Documents from
the Bar Kokhba Period in the Cave of Letters II: Hebrew, Aramaic and Nabataean Documents (),
–.
. Pliny, NH , ; Josephus, Ant. , ; Eusebius, Onom., ed. Klosterman, , .
. Acts :; Josephus, BJ , .

Rome and the East
large, since both Latin (in certain specific contexts) and—much more generally—Greek functioned as the languages of the dominant power, we are
dealing with the transmission of Latin and Greek words and concepts into
Semitic languages rather than the other way round. The material need not be
surveyed again here.6 But it is relevant to note that transliterations of Latin
military and administrative terms into Semitic languages were generally derived not directly from Latin but indirectly through their Hellenised forms.
So legio becomes λεγεών in the New Testament (Mk. :) and (see below)
LGYWN’ in Palmyrene, while centurio appears as QṬRYWN’ in Palmyrene or
QNṬRYN’ in Nabataean.7 More significantly, a large range of Roman official
terms reappear in Semitic texts not as transliterated via Greek, but as transliterations of the corresponding Greek translations. A number of examples
will be given below.
The reverse process, the appearance of Semitic terms in Greek and Latin,
is far less common and is indeed largely confined to names: of places, persons,
and deities. Without using Semitic names in Latin and Greek transliteration
it would have been quite impossible for either society or government to have
functioned in this region at all. In that limited sense Semitic terminology
pervades the entire written material—literary texts, perishable documents,
and inscriptions—from the Roman Near East; detailed examples do not need
to be given. But it should perhaps be stressed that while the re-naming of
cities with Greek, or Latino-Greek names (‘‘Maximianopolis’’) was a process
which was continuous from Alexander into the late Empire, the overwhelming mass of villages retained their Semitic names; and a significant proportion
of these names have been retained continuously until today.8
In two specific contexts, however, Latin could be deployed in inscriptions
in a way which was not significantly different from that in which it functioned in other regions of the Empire. One was the Latin ‘‘island’’ formed
. D. Kennedy and D. Riley, Rome’s Desert Frontier from the Air (). Note also J. Wagner,
Die Römer an Euphrat und Tigris (Antike Welt, Sondernummer, ); Dizionario Epigrafico IV,  (), s.v. ‘‘Limes’’: –/ (Mesopotamia, Osrhoene, Syria, Judaea, Arabia);
A. Lewin, ‘‘Dall’Eufrate al Mar Rosso: Diocleziano, l’esercito e i confini tardo-antichi,’’ Athenaeum  (): . See M. G. Angeli Bertinelli, Nomenclatura pubblica e sacra di Roma nelle
epigrafi semitiche (); Schürer, Vermes, and Millar, History II, ff. (Greek and Latin loanwords in Hebrew and Aramaic); M. G. Angeli Bertinelli, ‘‘I Semiti e Roma: appunti di una
lettura di fonti semitiche,’’ in Serta Historia Antiqua I (), .
. See Angeli Bertinelli, Nomenclatura, s.v. ‘‘centurio’’ (Palmyrene); CIS II, no.  (Nabataean).
. See, e.g., E. Frézouls, ‘‘La toponymie de l’Orient syrien et l’apport des éléments macédoniens,’’ in La toponymie antique. Actes du colloque de Strasbourg  (), .
Latin Epigraphy

under Augustus by the foundation of the colonia of Berytus, with its extensive territory stretching over Mount Lebanon into the northern Bekaa
valley. Here the public use of Latin was to survive the creation by Severus
of a second colonia in this eastern sector, Heliopolis. But elsewhere, when
the creation of nominal coloniae became more frequent from the s onwards, Latin, in these ‘‘colonial’’ contexts, entered into complex relations
with Greek, and in some cases with Semitic languages.9
The other context for the relatively ‘‘pure’’ use of Latin was of course the
military one. Most notable of all are the Latin inscriptions of the stone-built
forts of the ‘‘desert frontier,’’ some still standing today, which line the edge of
the steppe from Mesopotamia to the Red Sea, accompanied by milestones,
above all those of the Via Nova in Arabia.10 The imperialist overtones of the
milestones of Trajan’s road are matched above all by the triumphant message
of the inscriptions from the Tetrarchic period. This appears most clearly in
the inscription from the ‘‘camp of Diocletian’’ at Palmyra:11
[Reparato]res orbi sui et propagatores generis humani dd.nn. Diocletianus [et Maximianus] invictis]simi impp. et Constantius et Maximianus nobb. Caess. castra feliciter condiderunt [curam age]nte Sossiano Hieroclete v(ir) p(erfectissimus), praess. provinciae, d(evoto)
n(umini) m(aiestati)q(ue) eorum.
But everywhere, in whatever terms military inscriptions proclaimed the
dominance of the emperors, individual soldiers lived in complex social,
economic, religious, and linguistic relations with the multi-lingual society
around them. This fact is very evident in the archive of Babatha (n.  above),
as it is also in the quite remarkable bilingual inscription of a Roman veteran
from the bank of the Tigris: the veteran (οὐετρανός), ‘‘Antonios Domittianos,’’ made his dedication to Zeus in Greek and Aramaic.12
It is displayed also, even more clearly, in the archive of Greek and Syriac
documents of the third century from the middle Euphrates.13 The region
. F. Millar, ‘‘The Roman Coloniae of the Near East: A Study of Cultural Relations,’’ in
H. Solin and M. Kajava, eds., Roman Eastern Policy and Other Studies (Societas Scientiarum
Fennica, Commentationes Humanarum Litterarum , ), , and esp. –, –, for
the Latin ‘‘island’’ of Berytus and Heliopolis ( chapter  of the present volume).
. For the ‘‘desert frontier,’’ see (if not very detailed as to inscriptions) the immensely
evocative book by Kennedy and Riley (n. ).
. J. Cantineau, Inventaire des inscriptions de Palmyre VI, no. .
. J. F. Healey and C. S. Lightfoot, ‘‘A Roman Veteran on the Tigris,’’ Epigraphica Anatolica
 (): .
. See D. Feissel and J. Gascou, ‘‘Documents d’archives romains inédits du Moyen

Rome and the East
to which these priceless new texts primarily refer, the area of the Euphrates around and upstream from its confluence with the Chabur, overlaps in
a very significant way with that covered in the very rich documentary material from Dura-Europos. This material too provides ample evidence of the
co-existence of Latin, Greek, and Semitic languages, and of the social, cultural, economic, and religious role of soldiers of the Roman army within a
civilian environment. It is only necessary to mention, by way of illustration,
the Greek deed of sale of .. , with signatures of witnesses in Latin, in
which a veteran of Cohors III Augusta Thracum purchases an orchard on the
banks of the Chabur. This papyrus, like the other documents on papyrus and
parchment, was included in a masterly volume of the Final Reports, published
in .14 But, in the context of a volume on Latin epigraphy, it is necessary to stress that the inscriptions of Dura, a unique combination of Latin,
Greek, Aramaic, Syriac, Safaitic, Palmyrene, and Middle Persian, have never
been revised or collected since their initial publication by Franz Cumont and
by others in the various reports of the excavations between the two world
wars.15 Nor does it seem that there are any plans to carry out this work. Yet
these texts, including graffiti and painted inscriptions, represent a unique
historical and linguistic resource which it remains very difficult to use; moreover they urgently require systematic revision, because many of the, often
hazardous, identifications of the buildings found by the excavators depend
on the relations of inscription and structure. Without a proper collection
and re-assessment of the inscriptions a critical evaluation of what we really
know of Dura as a city cannot be undertaken.16
Euphrate (IIIe s. après J.-C.) I. Les petitions (P.Euphr.  à ),’’ JS (): –; ‘‘II. Les actes
de ventes-achat (P.Euphr.  à ),’’ JS (): –; ‘‘III. Actes divers et letters (P.Euphr. 
à ),’’ JS (): –.
. C. Bradford Welles, R. O. Fink, and J. Frank Gilliam, eds., The Excavations at DuraEuropos, Final Report V.: The Parchments and Papyri (); the deed of sale of ..  is
no. .
. For the inscriptions of Dura, see F. Cumont, Fouilles de Doura-Europos (–)
(), –; Excavations at Dura-Europos, Preliminary Reports I–IX (–); Mesnil du
Buisson, Inventaire des inscriptions palmyréniennes de Doura-Europos (); R. N. Frye, J. F. Gilliam, H. Ingholt, and C. B. Welles, ‘‘Inscriptions from Dura-Europos,’’ YCS  (): .
Note also C. B. Welles, ‘‘The Population of Roman Dura,’’ in Studies in Roman Economic and
Social History in Honor of A. C. Johnson (), , and G. D. Kilpatrick, ‘‘Dura-Europos:
The Parchments and Papyri,’’ GRBS  ():  (the best available survey of linguistic
co-existence there).
. See F. Millar, ‘‘Dura-Europos under Parthian Rule,’’ in J. Wiesehöfer, ed., Das Parther-
Latin Epigraphy

Along with all these other aspects, the study of the role of Latin in Dura,
in public inscriptions and on perishable documents, and of Latin vocabulary
transliterated into other languages, remains to be undertaken. The situation
is at least better, however, in the case of the other major site in the Near East
for the study of linguistic co-existence and the influence of Latin: Palmyra.
Here at least there are systematic, if inevitably not complete, collections
of the inscriptions, as well as a brief but extremely penetrating modern account.17 The history of Palmyra cannot be reviewed in any detail here. But
three aspects will be considered, as a way of illustrating the complex questions which arise in relation to the role of publicly inscribed Latin in the
Near East: the significance of Palmyra as a new, and rapidly evolving, city;
the nature of the rather small corpus of inscriptions from Palmyra which are
trilingual, in Latin, Greek, and Palmyrene; and the way in which attention
to linguistic interplay between these three languages, as shown on contemporary inscriptions, can serve to correct mistaken views about the structure
of power in Palmyra in the third century.
Firstly, we need to recall the broad features of the history of Palmyra
which need to be borne in mind if the nature of its epigraphic record is to be
understood.18 As a city, Palmyra was not ancient, but came into existence in
the first century ..; only very slight evidence suggests the presence there
of tombs, and perhaps the original temple of Allat, in the second century ..
But although there had been no ‘‘city’’ of Palmyra in the Seleucid period,
dating by the Seleucid era, beginning in autumn  .., is universal in its
inscriptions. As a political community it evolved steadily towards the model
of the Greek city, acquiring a boulē and dēmos in the second half of the first
century .., along with a normal set of Greek city offices. In the course of
the first century it also came firmly within the orbit of the Roman province of Syria. But, unlike all other Greek cities in the Roman East, its people
continued to use a Semitic language, or dialect of Aramaic, Palmyrene, both
in communal, official inscriptions and in epitaphs. In the latter, Palmyrene
often appears alone, but in public inscriptions characteristically in a parallel
text with Greek. So far as our evidence goes, as an inscribed language, Palmyreich und seine Zeugnisse (Historia-Einzelschrift , ), – ( chapter  of the present volume).
. Corpus Inscriptionum Semiticarum II. (); J. Cantineau et al., Inventaire des inscriptions de Palmyre I–XII (–). See J. Starcky and M. Gawlikowski, Palmyre 2 ().
. This section sums up a more detailed treatment in The Roman Near East ( ..–..
) (), –.

Rome and the East
rene was also a novelty as used in the imperial period: the earliest known
text, from the temple of Bel, dates to  .. The ‘‘epigraphic habit’’ here was
a borrowing from the wider Greek world, but in a unique bilingual form.
The notion that under Hadrian Palmyra became a free city is a mistake,
although it was visited by the Emperor and acquired the Greek name ‘‘Hadrianic Palmyra’’ (HDRYN’ TDMR in Palmyrene), and it is under this name
that it appears in the famous bilingual customs law of the Seleucid year ,
.. .19 In the early third century, however, under Septimius Severus or
more probably Caracalla, it was granted the status of colonia, and it was as a
Roman colonia that it played its remarkable, but very brief, role in the troubled
events of the s and s.
It might have been expected that in the ‘‘colonial’’ period Latin, always
rare but attested on inscriptions from Palmyra from the reign of Tiberius onwards, would become much more common. In fact it does not, and indeed,
paradoxically, there are no Latin inscriptions of the colonia. Instead there is
an intensification of a linguistic phenomenon already visible in the ‘‘Greekcity’’ period: the Palmyrene vocabulary for Graeco-Roman city institutions
was almost entirely constructed by transliteration (for instance, in the customs law, DGM’, BWL’ WDMS, GRMṬWS, and NMWS’—nomos, the law
itself ). Attempts at translation are relatively rare, for instance RB ŠWQ—
‘‘controller of the market,’’ agaronomos. But this instance in fact comes from
the colonial period, when the new institutions of the colonia, clearly marked
in the inscriptions, are expressed characteristically in Greek or in Palmyrene
transliteration of Greek. Colonia indeed is expressed as QLNY’; but duumvir
as στρατηγός, and hence in transliteration as ’SṬRṬG.20 As in the Babatha archive (text to n.  above), Greek plays a dominant role in mediating between
Latin and a Semitic language.
Against that background, we may now turn to the quite brief corpus of
trilingual inscriptions from Palmyra, in which Latin, Greek, and Palmyrene
not merely influence each other but appear in formal parallel texts. In fact,
to my knowledge, there are only six such texts, dating between the s and
the s. As mentioned above, none is known from the colonial period. The
texts will be set out in chronological order.
. For the best treatment, see J. F. Matthews, ‘‘The Tax Law of Palmyra: Evidence for
Economic History in a City of the Roman East,’’ JRS  (): .
. See in more detail Millar (n. ), –.
Latin Epigraphy

Trilingual Inscriptions from Pre-colonial Palmyra
. M. Rodinson, ‘‘Une inscription trilingue de Palmyre,’’ Syria  ():
 (from just south of the agora).
.. 
Haeranes Bonne Rabbeli
f. Palmirenus phyles Mithenon
sibi et suis fecit.
Ἔτους γξτ´ μηνὸς Ξανδικοῦ
Αἰράνης Βωνναίου τοῦ Ῥαββήλου
Παλμυρηνὸς φυλῆς Μειθηνῶν ἑαυτῷ
καὶ Βωννῆ πατρὶ αὐτοῦ καὶ Βααλθηγα μητρὶ
αὐτοῦ εὐνοίας ἕνεκεν καὶ τοῖς ἰδίοις αὐτοῦ
BYRḤ NYSN ŠNT CCCLXIII QBR’ DNH’ DY
ḤYRN BR BWN’ BR RB’L BR BWN’ BR ’TNTN BR
TYMY TDMRY’ DY MN PḤD BNY MYT’ DY BN’ ’L
BWN’ ’BWHY W‘L B‘LTG’ BRT BLŠWRY DY MN
PḤD BNY GDYBWL ’MH WLH WLBNWHY LYQRHWN
Translation of Palmyrene:
In the month of Nisan of the year , this is the tomb of Hairan, son
of Bonne, son of Rabbel, son of Bonne, son of Athenatan, son of
Taimai, a Tadmorean from the tribe of the sons of Mita, which he
built for Bonne his father and for Baalthega, daughter of Bolsari
from the tribe of the sons of Gadibol, his mother, for himself and
for his sons, in their honour.
. M. Gawlikowski, ‘‘Deux publicains et leur tombeau,’’ Syria  (): 
(from the western entrance to the Valley of the Tombs).
.. /
[C. Virius Alcimus et]
[T. Stat]ilius Her[mes]
[fec]erunt sibi et suis
Ἔτους ηξτ´
[Γαίο]ς ᾽Υίριος Ἀλκίμος
[καὶ Τίτο]ς Στατίλιος Ἓρμης
[ἐποί ]ησαν ἐαυτοῖς καὶ τοῖς ἰδιοῖς

Rome and the East
GYS WYRS ‘LQMS WṬYṬS W’SṬṬLYS H[RMS]
BNW NPŠH WM‘RTH ‘LM LHN WLBNYHWN LYQ[RHWN]
BŠNT 
Translation of Palmyrene:
Gaius Virius Alcimus and Titus Statilius Hermes have constructed
this funerary monument and hypogaeum, for ever, for themselves
and their sons in their honour, year .
. CIS II, no.  IGR III, no.  J. Cantineau, Inventaire des
inscriptions de Palmyre VIII, no.  (found in the Temple of Bel, but
evidently deriving from a tomb).
.. 
[L.S]pedius Chrysanthus
[vi]vos fecit sibi et suis
Λούκιος Σπέδιος Χρύσανθο[ς]
ζῶν ἐποίησεν ἑαυτῷ καὶ τ[οῖς]
ἰδ[ίοι]ς ἔτους θξτ´ μηνὸς Γ [ορ]π[ιαίου].
BYRḤ ’LWL ŠNT CCCLXIX [BN]H [LWQY]WS
’SPDY[S] KRYSTWS MKS’ BḤYWHY [QBR’ DNH]
LH WLBNWH’ WLBNY BYTH LY[Q]RH[WN]
Translation of Palmyrene:
In the month of Elul of the year , Lucius Spedius Christos, tax
collector, during his life built this tomb for himself and his sons
and the sons of his house, in their honour.
. J. Cantineau, ‘‘Tadmorea,’’ Syria  (): , on ff. (a column from
the temple of Bel).
.. 
Bu[le et civi]tas Palmyrenorum Hairanem
Bo[nnae f.] qui et Rabbilum
pium [et phi]lopatrin
Ἡ [ βουλ]ὴ καὶ ὁ [δῆμος] Αἱράνην Βωννέο[υς]
[τὸν καὶ Ῥ ]άββηλο[ν]
κ[οσμητ]ὴν εὐσε[ βῆ] καὶ [φιλ]όπατριν, τειμῆς χάριν
[ἕ ]τους ἐπτ´ μηνὸς Ξανδικοῦ.
Latin Epigraphy

BWL’ WDMS LH[Y]RN BR BWN[’ DY MTQR’ RB’L]
MṢBTH BBNYBY [’]L[HY’] WRḤYM
MDYNTH ’QYMW LH ṢLM’ DNH LYQRH
BYRḤ N[YS]N [ŠNT] [CCC]LXXXV
Translation of Palmyrene:
The boule and demos, to Hairan, son of Bonne, who is (also) called
Rabbel, decorator of the buildings of the gods(?), and lover of his
city, erected to him this statue, to honour him, in the month
Nisan of the year .
. H. Seyrig, ‘‘Inscriptions grecques de l’agora de Palmyre,’’ Syria  ():
, no. ; Cantineau, Inventaire X, no.  (from just to the south of
the agora).
.. 
L. Antonio Callisstrato manc. IIII merc.
Galenus actor
Λ. Ἀντωνίῳ Καλλιστράτῳ τεταρτώνῃ Γαληνὸς πραγματευτὴ[ς] ἴδιο[ς].
ṢLM’ DNH DY (L)WQYS ’NṬWNYS
QLSṬRṬS DY RB‘ ’ DY
’QYM LH LYQRH GLNWS PRGMṬT’ DYDH BYRḤ ’(B?) ŠNT
CCCCLXXXV.
Translation of Palmyrene:
This is the statue of Lucius Antonius Callistratus ‘‘of the quarter,’’
which Galenus, pragmateutes, his agent, erected to him, to honour
him, in the month Ab(?) of the year .
. Kh. Al-As‘ad and J. Teixidor, ‘‘Quatres inscriptions palmyréniennes
inédites,’’ Syria  (): , no.  (find-spot not recorded; Museum of
Palmyra).
.. 

Rome and the East
[ἔκτι]σεν Λούκιος
[. . .]λῷου τοῦ ζπυ´
[. . du]arum gub
[. . pecun]ia sua
[Polli]one II et Apro II cos.
QBY’ ’LN TR[Y’]
’NṬNYS Q[LSṬRṬS]
The issues raised by these six inscriptions are of remarkable interest and complexity. In the earliest ( and ), although the Latin comes first and the Greek
second, the Palmyrene text is considerably fuller, and the others are summaries of it. Inscription  certainly comes from a tomb, and in it the prominent Palmyrene Hairan (who also reappears in ) lists his forefathers for five
generations; the earliest, Taimai, ought thus to have been born at some time
around the end of the second century .., that is the earliest period to which
the evidence relating to Palmyra takes us back. These male ancestors evidently belonged to the ‘‘tribe’’ of the ‘‘sons of Mita,’’ while Hairan’s mother
came from the ‘‘sons of Gadibol’’; the nature of these groupings is a complex question which will not be pursued here.21 From the epigraphic point of
view it is more significant that the Greek supplies a translation for TDMRY’
DY MN PḤD BNY MYT’, namely Παλμυρηνὸς φυλῆς Μειθηνῶν, and that
this translation is then transliterated, without alteration of grammar, into
the Latin version. This is the earliest attested use of Latin by a native Palmyrene, and might be taken as a piece of exhibitionism by a prominent citizen
in the early period of Roman domination.
Inscription  is also from a tomb. But in this case it is quite uncertain
whether it reflects a degree of attachment to Graeco-Roman culture by a
native Palmyrene or the presence of an immigrant Greek who is already in
possession of the Roman citizenship. If he is a Palmyrene, these names are
unique there,22 and he also claims no place in Palmyrene society, nor, in the
Greek and Latin texts, any specific position. But in the slightly fuller Palmyrene text he records a position described as MKS’, which in the customs law
appears in the Palmyrene text as the equivalent of ‘‘whoever has contracted
for the collection [of the tolls]’’ in the Greek text. It does not however follow that he is a Roman publicanus. The contractors mentioned in the customs
law are collecting tolls for the city, even if under the terms of regulations
. See D. Schlumberger, ‘‘Les quatre tribus de Palmyre,’’ Syria  (): , and cf.
M. Gawlikowski, Le temple palmyrénien (), chap. .
. See J. K. Starck, Personal Names in Palmyrene Inscriptions ().
Latin Epigraphy

made under Roman supervision. It is equally possible that this tax collector
is an immigrant Greek, with Roman citizenship, who is acting for the city.
One may note as a parallel from the second century a bouleutēs from Antioch,
also a Roman citizen, who was τεταρτώνης (collector of a  percent tax),
evidently at Palmyra. He is recorded on a bilingual inscription of .. .23
In either case the significance of inscription  as a sign of the way in
which Palmyra was being rapidly drawn into the orbit of Graeco-Roman
culture is far outweighed by that of , which represents the earliest appearance of the boulē and dēmos of Palmyra. Unlike the earlier three, this is the
base of a statue (ṢLM’—‘‘image’’), erected to the same Hairan as in inscription  by the (as it seems) newly formed, Greek-style deliberative bodies of
the city. Given difficulties with the text, whether he is being credited with
contributing to the building, or decoration, of temples remains uncertain;
but the notion of being philopatris (RḤYM MDYNTH) is present in all these
texts. It is of some significance that the Palmyrene equivalent is a translation,
whereas that in the Latin—[phi]lopatrin—is a transliteration (not, alas, to be
found in the Oxford Latin Dictionary). By contrast, the standard transliteration
BWL’ WDMS now takes its established place in Palmyrene public inscriptions, while the Latin equivalent is an awkward mixture of transliteration
and translation, ‘‘bu[le et civi]tas Palmyrenorum.’’ As so often, both versions
are derivatives of the Greek.
Inscriptions  and  evidently concern the same man, L. Antonius Callistratus. The fact that he is a Roman citizen of Greek origin, given the dating
of the inscription in the s, should cause no surprise. He too, like Spedius
Chrysanthus, is a tax collector, concerned with a  percent tax, apparently
on goods being traded or transported. This role is described in markedly different ways in the three languages: τεταρτώνης in Greek, and DY RB’ in
Palmyrene (as in the inscription of  mentioned above); both thus give the
level of the tax but do not indicate its object (and again the Palmyrene can
hardly have been intelligible without the Greek). In Latin we have ‘‘manceps
of the fourth part[?] of the mercatura, or mercatus or mercedes[?].’’ Again it is
clear that the concepts πραγματευτής in Greek and actor in Latin are independent equivalents; but the Greek goes into Palmyrene in transliteration,
PRGMṬT’—but is helped out by a Palmyrene gloss, DYDH, literally ‘‘who
[is at] his hand.’’
Inscription  is conspicuously ‘‘Roman,’’ in giving, as well as the Seleucid year in Greek, a dating in Latin by the consuls of , a feature which
would not reappear in the inscriptions of Palmyra as a colonia. But it is also
. J. Cantineau, Inventaire X, no. . τεταρτώνης here is translated as DY RB‘ ’.

Rome and the East
striking in that, fragmentary as it is, it seems to exhibit the transliteration
of a Palmyrene word into Latin. For GUB in the Latin text seems to be a
transliteration of GBY’, possibly meaning ‘‘craters,’’ which may have been recorded as offerings. With this fragmentary text, in which for the first time
the Latin appears between the Greek and the Palmyrene, instead of first, the
Latin epigraphy of the city of Palmyra comes to an end, a few decades before
it rose to colonial status and acquired a new constitution. The only exceptions are a military inscription of .. , and then, after the collapse of
Palmyrene power in the s, inscriptions of the Tetrarchic period. But it was
in the context of the colonial period that one family at least rose to Roman
senatorial rank.
The Position of Septimius Odenathus
and His Son Septimius Hairanes in the s
The rest of this paper will not be concerned with the dramatic events of
the s and s which briefly brought a prominent Palmyrene, Septimius
Odenathus, and then his widow Zenobia and his son Septimius Vabalathus,
to a leading role in the history of the Roman Near East.24 Instead it builds on
the re-interpretation of the presumed earlier history of the dynasty offered
in a very important article by Michal Gawlikowski, published in .25 Here
Gawlikowski showed that we have no evidence from before the s for
members of the family occupying a leading role in Palmyra (or, still less, as
representing a line of local dynasts). What we have instead, in the s, is a
picture of an important local family, of Roman senatorial status, whose position is reflected in a series of honorific inscriptions in Greek and Palmyrene.
Roman concepts, and individual Latin words, are everywhere present. But
the Latin language, as such, is not. None the less, a study of the role of Latin
terminology in the Greek and Palmyrene public vocabulary of the colonia in
this period can add considerably to the force of Gawlikowski’s argument.
The claim to imperial power by the family was to be a product of later circumstances, and was made by a senatorial family from within the context
of the Roman colonia from which they came. What that context was is the
subject of this section. The texts need again to be set out in order. It should
be stressed that these are the earliest dated inscriptions put up to honour any
. On the s and s, see F. Millar, ‘‘Paul of Samosata, Zenobia and Aurelian: The
Church, Local Culture and Political Allegiance in Third-Century Syria,’’ JRS  (): –
( chapter  of the present volume).
. M. Gawlikowski, ‘‘Les princes de Palmyre,’’ Syria  (): .
Latin Epigraphy

members of the family. The three bilingual inscriptions come from the Great
Colonnade, and the three Greek ones from the temple of Baalshamin.
. CIS II., no.  OGIS, no.  Cantineau, Inventaire III, no.  Gawlikowski, Syria  (): no. .
October, .. 
Σεπτίμιον Αἱράνην Ὀδαινάθου, τὸν λαμπρότατον συνκλητικόν,
ἔξα[ ρχον? τε Παλμυ]ρηνῶν,
Αὐρήλι[ος Φιλῖνο]ς Αὐρ. Ἡλιοδώρου [τοῦ Ῥααίου] στρατιώτης λεγ[εῶνος Κυρηνα]ϊκῆς, τὸν
πάτρωνα, τειμῆς καὶ εὐχαριστίας χάριν, ἕτους γξφ´.
line : or
[.. Φλαβιανό]ς
ṢLM’ DNH DY SPṬMYWS ḤYRN BR
’DYNT SNQLṬYQ’ NHYR’ WRŠ
TDMWR DY ’QYM LH ’WRLYS
PLYNOS BR MRY’ PLYN’ R’Y PLḤ’
DBLGYWN’ DY BṢR’ LYQRH BYRḤ
TŠRY DY ŠNT DLXIII
Translation of Palmyrene:
This is a statue of Septimius Hairan, son of Odenathus, most
distinguished senator and head of Tadmor, which Aurelius
Philinus [or Flavianus] son of MRY’ Philinus [or Flavianus] R’Y, a
soldier who (is) in the legion which (is at) Bostra, erected for him,
in his honour. In the month of Tishri of the year .
. Gawlikowski, Syria  (): no. . Published by Kh. As‘ad and
M. Gawlikowski, ‘‘New Honorific Inscriptions in the Great Colonnade
of Palmyra,’’ AAAS – (–): –, no. .
April, .. 
Σεπτίμιον Ὀδαίνα[θον Αἱ ] ράνου Ὀ[υαβ ]αλλάθ[ου τοῦ Νασώρου] λα[ μ]πρό-

Rome and the East
τατον [ἔξαρχον Παλμυ]ρηνῶν, Ἰουλίος Αὐρήλιος
Ἀθηακά[ βος Ὀ]γήλου Ζαβδιβώ[λου ..] ΚWΣΣΚ τοῦ καὶ Κω( ρρα ...)
τατον φίλον στοργὴς ἕν (εκεν)
ἕτους γξφ´ μηνεὶ
[Ξανδ ]ικῷ.
ṢLM ’SPṬMYWS ’[DYNT BR ḤYRN] BR WHBLT NṢWR RŠ[’] DY
[TDMW]R NHYR’ D‘BD LH ’T‘QB BR ‘GYLW BR ZBDBWL BR
MQYMW
DMQR’ QR’ RḤMH LYQRHWN BRBNWTH BYRḤ NYSN ŠNT
DLXIII
Translation of Palmyrene:
Statue of Septimius O[denathus son of Hairan], son of Vaballathus
Nasor, distinguished head of Tadmor, which Atheacabus son of
Ogelus, son of Zabdibolus, son of Mocimus, who is (also) called
Cora, his friend, made for him, in their honour, in his presidency.
In the month of Nisan of the year .
. CIS II., no.  Cantineau, Inventaire III, no.  Gawlikowski,
Syria  (): no. .
April, .. 
Σεπ[τίμιον Ὀδαίναθον]
τὸν λαμ[πρότατον ὑπατικ]ὸν
συντέ [λεια or – χνια τῶν χρυσοχ ]όων
καὶ ἀργυ[ ροκόπων τ]ὸν δεσπότην
τειμῆς χάριν, [ἕτου]ς θξφ´
μηνεὶ Ξανδικῷ.
ṢLM’ SPṬMYWS ’DYNT
NHYR’ HPṬYQ’ MRN DY
’QYM LH TGM’ DY QYNY’
‘BD’ DHB’ WKSP’ LYQRH
BYRḤ NYSN DY ŠNT DLXVIIII
Translation of Palmyrene:
Statue of Septimius Odenathus, most distinguished consularis, our
master, which the association [tagma] of workers in gold and
Latin Epigraphy

silver, erected to him, in his honour. In the month Nisan of the
year .
. Chr. Dunant, Le sanctuaire de Baalshamin à Palmyre III: les inscriptions
(), no.  Gawlikowski, Syria  (): no. .
.. /
Σεπτιμίον Ὀδαίναθον
τὸν λαμπρότατον ὑπατικὸν
τὸ συμπόσιον τῶν κονετων
τὸν πάτρωνα
ἕτους θξφ´.
. Dunant, Le sanctuaire de Baalshamin à Palmyre III: les inscriptions (), ,
n.  Gawlikowski, Syria  (): no. .
.. /
[Σεπτίμι]ον Ὀδαίναθον
[τὸν] λαμπρότατον
[ὑπ]ατικὸν
[τὸ] συμπόσιον
[τῶν] Ουαννων τὸν
[π]άτρωνα
ἕτους θξφ´
. H. Seyrig, ‘‘Les fils du roi Odainat,’’ Ann. Arch. Ar. Syr.  ():  Gawlikowski, Syria  (): no. .
.. /
[Σεπτίμιον] Αἱράνην τὸν
λαμπρότατον (υἱὸν) Ὀδαινάθου τοῦ λαμπροτάτου ὑπατικοῦ τὸ
συμπόσιον σκυτο- (sic)
ων καὶ ἀσκοναυτο –
ποιῶν τὸν πάτρωνα
ἕτους θξφ.´
The two inscriptions of the Seleucid year  (.. /) relate to a son of
Odenathus, Septimius Hairanes, who is himself already of Roman senatorial
rank, συγκλητικός/SNQLṬYQ’. It is extremely significant that the earliest
of all (no. ) is an honorific inscription put up to him not by the Palmyrenes

Rome and the East
but by a soldier from the Legion III Cyrenaica, stationed at Bostra: in Palmyrene PLḤ’ [soldier] DB LGYWN’ DY BṢR’. Septimius Hairanes was his
patronus, duly transliterated as πάτρων. This expression finds no equivalent
in the Palmyrene version.
By contrast, the Palmyrene text of inscription , like that of inscription ,
does give Hairanes a title which might well suggest an established dynastic
position: RŠ TDMWR or RŠ [’] DY [TDMWR], ‘‘head of Tadmor.’’ But we
have to be careful in attributing too definite a meaning to this expression,
which is not found elsewhere in the epigraphy of Palmyra. Moreover, an
Aramaic and Greek bilingual inscription of the second–third century, from
Khirbet Zif in ancient Idumaea, uses RŠ plus an expression denoting either a
group of people or a locality as the equivalent of the Greek πρωτοπολείτης,
‘‘leading citizen.’’ 26 Is there any reason to suppose that the soldier from Bostra
meant anything more precise than that? The Greek equivalent in this case is
baffling, partially missing in inscription  (ἔξα[ ρχον τε Παλμυ]ρηνῶν) and
almost completely missing in inscription  ([ἔξαρχον Παλμυ]ρηνῶν). It is
clear that Hairanes is described as something ‘‘of the Palmyrenes’’; but what?
Restoration depends entirely on the three letters ἔξα[. . .]. The normal restoration, ἔξα[ ρχον], may well be correct, though it requires inserting τε to
fill up the space. It might be equally possible to restore ἔξα[ιρετός]. But
neither expression has any established usage as a technical term or in honorific inscriptions (Plutarch, however, uses ἔξαρχος τῶν ἱερέων for pontifex
maximus).27 Even without the parallel from Khirbet Zif it should be clear
that we cannot base any conception of a specific role on these two restored
expressions. The context is in any case firmly Roman.
All the remaining inscriptions of the s date to the Seleucid year ,
.. /, that is, while Valerian was in Antioch, and not long before his
fatal campaign into Mesopotamia. The only evidence that an independent
military role on the part of Odenathus had begun before this is provided
by Malalas, who records that he attacked and pursued the retreating Persian
forces in ; but he has very probably retrojected a role which Odenathus
in fact played only in the s.28
At any rate the honorific inscription relating to Septimius Hairanes (no. )
. L. Y. Rahmani, ‘‘A Bilingual Ossuary: Inscription from Khirbet Zif,’’ IEJ  ():
 J. A. Fitzmyer and D. J. Harrington, A Manual of Palestinian Aramaic Texts (), ,
no. A. See also Y. Yadin, ‘‘A Note on the Bilingual Ossuary Inscription from Khirbet Zif,’’
IEJ  (): –.
. Plutarch, Numa .
. See Millar (n. ), –.
Latin Epigraphy

now no longer uses the term ἔξα[ ρχος?] Παλμυρηνῶν, but refers strictly
to his Roman status—λαμπρότατος (vir clarissimus, that is a senator)—and
son of Odenathus, the lamprotatos hypatikos. This too is, of course, a Roman
status, clarissimus consularis. It would be of considerable historical importance
to know if consularis/hypatikos is being used here in the sense of ‘‘governor’’—
which would imply that he was governing the (praetorian) province of Syria
Phoenice.29 The alternative is that this is simply a status designation. But the
question may remain undecided in this context. The essential point is that
this is indeed a Roman status designation and that the context of this honorific inscription, as of the three erected to Odenathus himself, is civilian.
All four are erected by trade or professional groups within Palmyra. These
themselves present problems which need not be discussed here.
What light do these inscriptions throw on Odenathus’ position within
the structure of the Palmyrene community in the late s? We may begin
with the bilingual inscription of the association of gold- and silver-workers
(no. ). Interestingly, to describe their association they use a different transliterated Greek word in the Palmyrene text from that employed in the Greek
text: TGM’ (τάγμα) DY QYNY’ ‘BD’ DHB’ WKSP’. The Greek text, largely
restored, gives Odenathus, as well as his Roman rank, consularis, the appellation δεσπότης, ‘‘master,’’ almost precisely reflected in the Palmyrene text,
which calls him MRN, ‘‘our master.’’ So is this perhaps a reference to his
dynastic role in Palmyra? The Greek inscriptions from the temple of Baalshamin, one in honour of Odenathus’ son Hairanes (no. ), and two for Odenathus himself (nos.  and ), show that any such notion would be completely misconceived. For, like the inscription put up some years earlier by
the soldier of III Cyrenaica, these inscriptions reflect a personal relationship,
in these cases between associations and a prominent citizen. The meaning
which should be given to δεσπότης and MRN in inscription , set up by one
trade-association, is clearly indicated by the term which is used in the three
other contemporary inscriptions of trade associations (nos. –), as it had
been in that from the soldier: πάτρων. Far from suggesting that Odenathus
was already a ‘‘prince’’ in Palmyra, these texts illustrate how profoundly Palmyra, while still—uniquely—retaining the public use of a Semitic language,
had taken on the character of a Graeco-Roman city; and more specifically
how, as a Roman colonia whose leading family now enjoyed high senatorial
rank, Roman concepts of status and relationship had come to play a part in
its social structure.
. See B. Rémy, ‘‘ Ὑπατικοί et consulares dans les provinces impériales prétoriennes
au IIe et IIIe siècles,’’ Latomus  (): .

Rome and the East
These two brief sketches of two aspects of the complex linguistic and epigraphic map of the Roman Near East can be no more than suggestive of the
riches which remain still largely unexplored. The epigraphic use of Aramaic
and its dialects evolved, throughout the region, in step with that of Greek,
which was always the dominant language, and with that of Latin. If Latin
always played a smaller role, it was not a negligible one, if only because all
of the Near East was at all times a military area, in which the influence of
the Roman army was profound. But for us to understand these mutual social, cultural, and linguistic influences more fully, the most urgent task is the
reactivation of the epigraphic heritage of Dura-Europos.
 
Paul of Samosata, Zenobia, and Aurelian:
The Church, Local Culture, and Political
Allegiance in Third-Century Syria *
Introduction
What we call the ‘‘eastern frontier’’ of the Roman Empire was a thing of shadows, which reflected the diplomatic convenience of a given moment, and
dictated the positioning of some soldiers and customs officials, but hardly
affected the attitudes or the movements of the people on either side.1 Nothing more than the raids of desert nomads,2 for instance, hindered the endless
movement of persons and ideas between Judaea and the Babylonian Jewish
community.3 Similarly, as Lucian testifies, offerings came to the temple of
Atargatis at Hierapolis/Bambyce from a wide area of the Near and Middle
* Originally published in JRS  (): –. I was very grateful, for discussion and correction, to Professor A. D. Momigliano, Professor G. D. Kilpatrick, Professor P. R. L. Brown,
Professor P. J. Parsons, Dr. J. Rea, Professor T. D. Barnes, and especially to M. Henri Seyrig.
. See Philostratus, Vit. Ap. Ty. , , for Apollonius’ famous confrontation with the
customs official at Zeugma. The only evidence known to me of the frontier actually preventing movement comes in Jerome, Vita Malchi  (PL XXIII, ), where Malchus, from
Nisibis, relates that (sometime in the first half of the fourth century) ‘‘because I could not
go to the east, because of neighbouring Persia and the Roman military guard, I turned my
feet to the west.’’
. Note Herod’s establishment of a colony of Babylonian Jews in Batanea for the protection of caravans of pilgrims coming from Babylonia to Jerusalem. Jos., Ant. , –;
Vita, –.
. For visitors from Mesopotamia, see J. Jeremias, Jerusalem in the Time of Jesus (),
–, and for the cultural and personal relations of the two communities the successive
volumes by J. Neusner, A History of the Jews in Babylonia I: The Parthian Period 2 (); II: The
Early Sasanian Period (); III: From Shapur I to Shapur II (); IV: The Age of Shapur II
(); V: Later Sasanian Times ().


Rome and the East
East, including Babylonia.4 The actual movement to and fro of individuals
was reflected, as we have been reminded,5 in a close interrelation of artistic
and architectural styles. Moreover, whatever qualifications have to be made
in regard to specific places, it is incontestable that Semitic languages, primarily Aramaic in its various dialects, remained in active use, in a varying
relationship to Greek, from the Tigris through the Fertile Crescent to the
Phoenician coast. This region remained, we must now realize, a cultural
unity, substantially unaffected by the empires of Rome or of Parthia or Sassanid Persia.6
On the face of it, these facts might seem to give added confirmation to
what is the standard interpretation of the career of Paul of Samosata. The
bare structure of this career, which we know essentially from Eusebius,7 is
that he succeeded Demetrianus as bishop of Antioch, was accused of heresy,
and was the subject of two synods held at Antioch in about  and / (for
the dates, see below), the second of which condemned him. On his refusal to
leave the church house, his opponents made a successful petition against him
to Aurelian (–). The accepted interpretation, represented primarily by
G. Bardy’s book on Paul,8 and by Glanville Downey’s standard work on the
history of Antioch,9 sets this career firmly in the context of a wide pattern of
cultural and political relationships. On this view, Paul, coming from Samosata, was the champion of the ‘‘native’’ (Syriac- or Aramaic-speaking) element in the Antiochene church. His opponents were the representatives of
Greek culture. Moreover, a remark made in the letter of the second synod
retailing Paul’s offences is held to mean that he held a government post as
ducenarius (i.e., a procurator with a salary of , sesterces per annum; see
text to n.  below); and later evidence is brought in to show that this will
have been in the service of Palmyra, more specifically of its queen, Zenobia.
Thus a conflict of cultures becomes intimately linked to a political conflict.
. Lucian, dea Syra , . See text to n.  below.
. J. B. Ward-Perkins, ‘‘The Roman West and the Parthian East,’’ Proc. Brit. Acad.  ():
; ‘‘Frontiere politiche e frontiere culturali,’’ La Persia e il mondo greco-romano, Acc. Naz.
dei Lincei, anno , quad.  (): .
. See the remarks by P. Brown, ‘‘The Diffusion of Manichaeism in the Roman Empire,’’
JRS  (): .
. Eusebius, HE , –.
. G. Bardy, Paul de Samosate: étude historique 2, Spicilegium Sacrum Lovaniense IV ();
note, however, the more cautious view of Paul in G. Bardy, La question des langues dans l’église
ancienne (), .
. G. Downey, A History of Antioch in Syria from Seleucus to the Arab Conquest (), –
, –.
Paul of Samosata

Paul owes his position to Zenobia, and his opponents’ appeal to Aurelian will
have taken place when the latter recaptured Antioch from the Palmyrenes
in . In spite of the partial reservations in the excellent earlier study of
Paul by F. Loofs 10 and the briefly expressed scepticism of A. Alföldi,11 this
interpretation seems to have become firmly established.12 It also serves what
appears to be the necessary function of explaining how the synod of Antioch could have petitioned a pagan emperor, and how that Emperor found it
worthwhile to attend to their request, and to give it a favourable response.
The career of Paul of Samosata is thus central to a number of questions of
great delicacy and importance. The collection and arrangement of the scanty
and disparate evidence for the various elements of a possible ‘‘local’’ culture
in any area of the Roman Empire are difficult enough; much more sensitive
is the question of the role, function, or prestige of that culture in relation to
a dominant culture.13 It is a still more hazardous step to assert that what we
know of any individual episode justifies the imposition of an explanation in
terms of a local ‘‘nationalism.’’ 14 But at the same time the very fragility and
scantiness of our evidence is itself a reason for not proceeding with brusque
confidence to negative conclusions.
Syriac and Greek in the East Syrian Regions and Mesopotamia
From scattered evidence we can now gain some conception of the geographical spread and profound influence of Greek culture through Meso. F. Loofs, Paulus von Samosata; eine Untersuchung zur altchristlichen Literatur und Dogmengeschichte, Texte und Untersuchungen XLIV,  (), esp. .
. CAH XII, , n. : ‘‘The political connections of Zenobia with Bishop Paul of Antioch seem to the present writer even less real than to Fr. Loofs.’’ For the relevance of Alföldi’s
classic studies of the coinage in this period, see below text to nn. – and notes.
. See, e.g., F. Caspar, Geschichte des Papsttums I (), ; J. Lebreton and J. Zeiller,
Histoire de l’Église II (), ; H. Grégoire, Les persécutions dans l’Empire romain 2 (),
; J. Daniélou and H. Marrou, Nouvelle histoire de l’Église I: des origines à Saint Grégoire le
Grand (), ; W. H. C. Frend, Martyrdom and Persecution (), –; B. Altaner and
A. Stuiber, Patrologie 7 (), ; H. Chadwick, The Early Church (), –.
. For the parallel case of North Africa, see the contrasted treatments by F. Millar,
‘‘Local Cultures in the Roman Empire: Libyan, Punic and Latin in Roman Africa,’’ JRS 
():  ( chapter  of F. Millar, Rome, the Greek World, and the East II: Government,
Society, and Culture in the Roman Empire), and P. Brown, ‘‘Christianity and Local Culture in
Late Roman Africa,’’ JRS  (): .
. For a cautious and useful survey of this question in another region, see R. MacMullen, ‘‘Nationalism in Roman Egypt,’’ Aegyptus  (): .

Rome and the East
potamia to Iran, central Asia, and Afghanistan.15 By contrast, we have from
Judaea the example of a society close to the Mediterranean seaboard, whose
culture was very early deeply affected by Hellenism,16 and where the speaking of Greek was clearly widespread,17 but which consciously maintained
and developed a local culture and tradition within its Hellenised environment. We could hardly ask for a neater example of the conflict of cultures
than the story of R. Gamaliel II elaborately justifying his use of the baths at
Akko-Ptolemais, in spite of the presence there of a statue of Aphrodite.18
Moreover, we have at least one case where a change of political regime
does seem to be immediately reflected in the predominance of Greek as an
official language. In both language and art the kingdom of Nabataea, annexed in ,19 belonged to the Aramaic-Greek world mentioned above. But
while the Nabataean language, as is shown by inscriptions,20 persisted at least
until the early fourth century, the ‘‘archive of Babatha,’’ discovered in ,
and now fully published, shows that to a significant extent Nabataean was
abandoned in legal documents and replaced by Greek within a few years of
the establishment of the province.21
. See the brilliant survey by E. Bickerman, ‘‘The Seleucids and the Achaemenids,’’ Persia e il mondo greco-romano (n. ), ; P. Bernard, ‘‘Aï Khanum on the Oxus: A Hellenistic City
in Central Asia,’’ Proc. Brit. Acad.  (): ; L. Robert, ‘‘De Delphes à l’Oxus. Inscriptions
grecques nouvelles de la Bactriane,’’ CRAI (): ; and in general D. Schlumberger,
L’Orient hellénisé ().
. See M. Hengel, Judentum und Hellenismus: Studien zu ihrer Begegnung unter besonderer
Berücksichtigung Palästinas bis zur Mitte des . Jh. v. Chr. (), and F. Millar, ‘‘The Background
to the Maccabean Revolution: Reflections on Martin Hengel’s ‘Judaism and Hellenism,’ ’’
Journal of Jewish Studies  (): – ( chapter  of the present volume).
. The evidence, from a variety of periods, is collected by J. N. Sevenster, Do You Know
Greek? How Much Greek Could the First Jewish Christians Have Known? ().
. Mishnah, Abodah Zarah, : (ed. Danby, ). On the more permissive attitude to
representational art which developed in the second and third centuries, see, e.g., C. H.
Kraeling, The Excavations at Dura Europos, Final Report VIII, I: The Synagogue (), –;
E. E. Urbach, ‘‘The Rabbinical Laws of Idolatry in the Second and Third Centuries in the
Light of Archaeological and Historical Facts,’’ IEJ  (): , .
. See G. W. Bowersock, ‘‘The Annexation and Initial Garrison of Arabia,’’ ZPE 
(): .
. See the excellent survey of J. Starcky, ‘‘Pétra et la Nabatène,’’ Dictionnaire de la Bible,
Supp. VII (), –, esp. .
. See N. Lewis, The Documents from the Bar Kokhba Period in the Cave of Letters: Greek
Papyri (); Y. Yadin, J. C. Greenfield, A. Yardeni, and B. Levine, The Documents from the
Bar Kokhba Period in the Cave of Letters: Hebrew, Aramaic and Nabatean Aramaic Papyri (),
Paul of Samosata

Thus the wider background provides an ample justification for posing
the question whether the development of Syriac literature, almost entirely
Christian, and of the distinctive Syriac script, reflects some wider cultural
and political movement, to which both the rise of Palmyra and heretical
movements in the church within Roman Syria might conceivably be related.
All the places which are most relevant—Osrhoene, Palmyra, and Samosata—belong to that wide region on either side of the Euphrates over which
Roman rule was steadily extended in the course of the first three centuries of
the Empire. Beyond this area too, in Parthian and Sassanid Mesopotamia and
Iran, it is generally accepted that cultural changes were taking place, with a
steady eclipse of Hellenism in the surviving Greek cities in the later Parthian
period;22 though it is notable that, as late as the end of the first century ..,
a Greek geographer could be produced by distant Charax at the head of the
Persian Gulf.23 Nor is it clear that it was mere vainglory that led to Greek
being one of the three languages of the great inscription of Shapur I on the
Kaaba of Zoraster at Naqsh-i-Rustam.24
The rise of Syriac belongs to the ‘‘frontier’’ area along the Euphrates and
to Osrhoene, with its capital Edessa. The Syriac cursive script is first attested
on an inscription of ..  from Birecik on the left bank of the Euphrates, and a couple of other inscriptions come from the same region later in
the century.25 More important is the earliest surviving Syriac document on
perishable material, the deed of sale written at Edessa in  and found at
Dura-Europos.26 From Edessa we have the apparently eye-witness account
of the flood of .. , later incorporated in the Syriac Chronicle of Edessa,27
with H. M. Cotton, ‘‘The Languages of the Legal and Administrative Documents from the
Judaean Desert,’’ ZPE  (): –.
. For the Greek colonies and cities of Mesopotamia, see, e.g., A. H. M. Jones, Cities
of the Eastern Roman Provinces 2 (), chap. IX; N. Pigulevskaja, Les villes de l’état iranien aux
époques parthe et sassanide (), esp. chaps. I–IV; M. A. R. Colledge, The Parthians (),
–.
. For the date of Isidorus of Charax, see S. A. Nodelman, ‘‘A Preliminary History of
Characene,’’ Berytus  (): , on –.
. For the text, A. Maricq, ‘‘Res Gestae Divi Shaporis,’’ Syria  (): .
. A. Maricq, ‘‘La plus ancienne inscription syriaque; celle de Birecik,’’ Syria  ():
; cf. J. Pirenne, ‘‘Aux origines de la graphie syriaque,’’ Syria  (): , and E. Jenni,
‘‘Die altsyrischen Inschriften, –. Jahrhundert n. Chr.,’’ Theologische Zeitschrift  (): .
. P. Dura , re-edited by J. A. Goldstein, ‘‘The Syriac Bill of Sale from Dura-Europos,’’
JNES  (): ; cf. H. M. Cotton, W. Cockle, and F. Millar, ‘‘The Papyrology of the
Roman Near East: A Survey,’’ JRS  (): no. .
. Ed. L. Hallier, Texte und Untersuchungen IX. (); see –.

Rome and the East
and more significantly the writings (all now lost except the Book of the Laws
of Countries) of the heretic Bardesanes (–c. ).28 In the face of this important development we perhaps forget that Edessa too was a Macedonian
colony.29 Bardesanes was literate in both Greek and Syriac,30 as was his son
Harmonius, who was educated in Athens, and composed hymns in Syriac;31
and Bardesanes’ Syriac works were translated into Greek by his disciples.32
In short, Bardesanes was the product of a mixed culture, where it may be
impossible for us to determine what the values attached to each language
were, or even, in certain cases, which the original language of a particular work was. Thus, for instance, only the most careful analysis can make it
probable that the Odes of Solomon were written in Syriac, and translated into
Greek by the third century, from when we have a papyrus text in Greek of
Ode XI.33
Edessa may not have been totally different from another Macedonian
foundation much better known to us, Dura-Europos. Here, together with
a preponderance of Semitic cults, Greek documents still far outnumber all
others (in Latin, Pahlavi and Middle Persian, Parthian, Safaitic, Palmyrene,
Aramaic, and Syriac), even from the latest, Roman, period of the city.34
Furthermore, one of the Greek documents (P. Dura ) serves to illustrate the
extraordinary tangle of confusions which tends to beset any attempt to portray the cultural framework of early Eastern Christianity. This is the Greek
fragment of the Diatessaron of Tatian; in spite of the obvious implications
of the name, and the fact that no source actually says so, it has frequently
been argued that this was originally composed in Syriac. In fact there are no
valid linguistic arguments against the prima facie deduction from this very
. See H. J. W. Drijvers, Bardaisan of Edessa ().
. See J. B. Segal, Edessa, ‘‘The Blessed City’’ (); see – for traces of Greek culture
there in this period.
. Epiphanius, Panarion , , ; cf. Theodoret, Haereticarum fabularum compendium , 
(PG LXXXIII, ), mentioning Syriac only.
. Sozomenus, Hist. Eccles. , , –; Theodoret (n. ).
. Euseb., HE , , ; Jerome, de vir. ill. .
. See J. A. Emerton, ‘‘Some Problems of Text and Language in the Odes of Solomon,’’
J. Theol. St.  (): .
. For surveys, see C. B. Welles, ‘‘The Population of Roman Dura,’’ in Studies in Roman
Economic and Social History in Honor of A. C. Johnson (), ; G. D. Kilpatrick, ‘‘DuraEuropos: The Parchments and the Papyri,’’ Greek, Roman and Byzantine Studies  ():
. See also F. Millar, ‘‘Dura-Europos under Parthian Rule,’’ in J. Wiesehöfer, ed., Das
Partherreich und seine Zeugnisse (Historia-Einzelschrift , ), – ( chapter  of
the present volume).
Paul of Samosata

early text that Greek was the original language.35 But what was the origin
of Tatian himself ? In his Address to the Greeks, itself written in Greek, he says
(claiming for Christianity the prevailing prestige of oriental wisdom,36 but
also neatly exhibiting the prevailing confusion of cultures): ‘‘These things, I,
Tatianos, arguing against the barbarians, composed for you, Greek men; although born in the Assyrian land, I was first educated in your [culture].’’ 37
‘‘Assyria,’’ it has been asserted,38 must thus refer to the land east of the Tigris.
But the word could be applied to southern Mesopotamia,39 and is even used,
by a literary conceit, of places within Roman Syria. Lucian, from Samosata,
calls himself an ‘‘Assyrian’’ (see below), while Philostratus seems to apply it to
the inhabitants of Antioch 40—and equally to a man from Nineveh.41 Clement, a near contemporary, calls Tatian simply, ‘‘the Syrian.’’ 42 Of all our sources
only the not always reliable Epiphanius says anything definite about his origins; after saying separately that he was ‘‘a Greek by birth’’ and that he was
(again) a ‘‘Syrian,’’ he records that his preaching began in Mesopotamia and
continued, after a visit to Rome, in the area of Antioch, Pisidia, and Cilicia.43 We cannot in fact state the origin of Tatian, any more than we can
of the ‘‘Assyrian’’ Prepon, who, according to the contemporary Hippolytus,
wrote against Bardesanes (whom he calls an ‘‘Armenian’’ 44—while Porphyry,
later in the century, calls him a ‘‘Babylonian’’).45 But the very fact that clear
definitions of locality and nationality are wanting has its own significance.
All that has been said applied equally to Paul’s native city, Samosata. Most
of what we know of it relates to the royal house finally deposed by Vespasian
about .. , with its mixed Iranian and Greek traditions, and the vast inscriptions in Greek relating to the royal cult, from the hierothesion (sacred
monument) of Mithridates Callinicus at Arsameia on the Nymphaios and
. See Kilpatrick (n. ), –.
. Compare A. J. Festugière, La révélation d’Hermés Trismégiste I (), chap. II, ‘‘Les
prophètes de l’Orient.’’
. Address to the Greeks, , ed. E. Schwartz, Texte und Untersuchungen IV. ().
. E.g., A. Vööbus, Early Versions of the New Testament: Manuscript Studies (), ;
P. Kahle, The Cairo Geniza 2 (), –.
. See A. Maricq, ‘‘La province d’Assyrie créée par Trajan,’’ Syria  (): ; cf. P.
Brown (n. ), .
. Philos, Vit. Ap. Ty. , .
. Philos, Vit. Ap. Ty. , .
. Strom. , /, ; see also Theodoret, Haer. fab. comp. ,  (PG LXXXIII, ).
. Panarion , .
. Hippolytus, Elenchus , , –.
. Porphyry, de abstinentia , .

Rome and the East
of Antiochus I on Nemrud Dagh, with its magnificent free-standing sculptures.46 We would expect that here, as in the other border regions absorbed
by Rome, Greek and a dialect of Aramaic persisted together. This is tentatively confirmed by the evidence of Lucian (who in different passages calls
himself both an ‘‘Assyrian’’ and a ‘‘Syrian’’)47 when in the Bis Accusatus (Twice
Accused)  he makes ‘‘Rhetoric,’’ in the role of his accuser, call him as a youth
‘‘a barbarian still, so far as he sounds, and still wearing a Persian garment after
the Assyrian fashion.’’ This could easily be dismissed as a depiction of standard
rhetorical exaggeration, referring at most to a local accent or a mere rusticity.
But we have just enough evidence to show that such a conclusion would be
over-hasty. For what may be the earliest product of Syriac literature, the letter of Mara bar Sarapion,48 is the work of a Samosatene writing to his son
in the period after the expulsion of some citizens from there (including the
writer) by the Romans; the occasion cannot be definitely determined, and
might be any moment from the capture in  to the third century. Moreover
we have an apparently genuine martyr act from Samosata, written in Syriac
and relating to the early fourth century.49
From the mere fact that Paul is described as a Samosatene we cannot
simply assume that he was born and brought up there.50 But if we grant that
it is probable that he had some substantial connection with the place, then
it is not unlikely that he spoke Syriac as well as Greek (for there can be no
possible doubt that he used Greek as bishop of Antioch). But, even so, the
speaking or writing of Syriac did not of itself represent a rival, ‘‘oriental,’’
culture. Just as early Syriac documents and literary works exhibit numerous
Greek loan-words,51 so for instance the letter of Mara bar Sarapion itself has
a Stoic colouring and is replete with allusions from the history of classical
Greece.52
Moreover, the same doubts which must be felt about the ‘‘orientalism’’
. See F. K. Dörner and R. Neumann, Forschungen in Kommagene (); F. K. Dörner and
T. Goell, Arsameia am Nymphaios: Die Ausgrabungen im Hierothesion des Mithridates Kallinikos
von – (). Cf. F. K. Dörner, Kommagene, ein wiederentdecktes Königreich 2 ().
. dea Syra : ‘‘and I write, being an Assyrian.’’ Scyth. : ‘‘us Syrians.’’
. Text and English translation (–) by W. Cureton, Spicilegium Syriacum ().
Cf. F. Schulthess, ‘‘Der Brief des Mara bar Sarapion,’’ ZDMG  (): ; R. Duval, La
littérature syriaque 2 (), –; A. Baumstark, Geschichte der syrischen Literatur (), .
. See Duval (n. ), .
. Compare the remarks of R. Syme, ‘‘Hadrian and Italica,’’ JRS  (): .
. See A. Schall, Studien über griechische Fremdwörter im Syrischen (), –.
. The writer uses, in the course of providing exempla, the names of Polycrates, Achilles,
Agamemnon, Priam, Archimedes, Socrates, Pythagoras, Palamedes, and Plato.
Paul of Samosata

of Paul apply to Palmyra itself. For while no one can question either the
total predominance of Semitic cults there or the vigour and splendour of the
native Palmyrene art and architecture, the city (now a Roman colony) was
officially bilingual in Palmyrene and Greek 53 (though Palmyrene nomenclature in particular suggests that the population was in fact largely of Arab
stock).54 But it is only the Historia Augusta which appears to imply that Zenobia could not write a letter in Greek.55 That point arises in connection with
one of the only two facts we can be said to know about the culture and historical outlook of Zenobia. The first is that she brought to her court Longinus, the foremost Greek literary scholar of his day.56 The second is that in
Egypt she identified herself with Cleopatra; so it was to her under this name
that a Greek rhetor, Callinicus of Petra, dedicated his history of Alexandria.57
Thus the wider cultural background of the third-century Near East is
fraught with ambiguities. Before coming to the events of the s and s
it remains to look at Roman Syria proper. Is there anything to suggest either
the survival of non-classical traditions or the development of a local strain
in Syrian Christianity?
Local Culture in Roman Syria
Abundant evidence illustrates the survival of pre-Hellenic cult centres in Roman Syria and its environs. One need only mention by way of example the
. For a survey, see J. Starcky, Palmyre (); cf. also le Comte du Mesnil du Buisson,
Les Tessères et les monnaies de Palmyre: un art, une culture et une philosophie grecs dans les moules
d’une cité et d’une religion sémitiques ().
. See A. Caquot, ‘‘Sur l’onomastique religieuse de Palmyre,’’ Syria  (): ; cf.
H. Seyrig in Syria  (): –.
. HA, Aurel. , : ‘‘It was regarded as a cruel thing that Longinus the philosopher
should have been among those killed. He, it is said, was employed by Zenobia as her teacher
in Greek letters, and Aurelian is said to have slain him because he was told that that overproud letter of hers was dictated in accordance with his counsel, although it was in fact
composed in the Syrian tongue’’ (Loeb translation).
. HA (n. ); Zosimus , , –; Photius, Bib. , . Bekker (see below text to
n. ); Syncellus ,  Bonn. For the text of a letter from Longinus written to Porphyry in
Sicily (so between c. , Porph., Vit. Plot. , and ) inviting him to join him in ‘‘Phoenicia,’’ see Porph., Vit. Plot. . Libanius, Ep.  Förster ( Wolf ), mentions a logos
Oudainathos of Longinus, presumably a funeral address. Cf. RE, s.v. ‘‘Longinus.’’
. See A. Stein, ‘‘Kallinikos von Petrai,’’ Hermes  (): ; J. Schwartz, ‘‘Les Palmyréniens en Égypte,’’ Bull. Soc. Ant. Alex.  (): ; A. D. E. Cameron in CQ, n.s., 
(): –.

Rome and the East
cult of Perasia at Hierapolis/Castabala 58 (in the province of Cilicia); of Atargatis at Hierapolis-Bambyce, brilliantly described by Lucian in the dea Syra,59
or of Elagabal at Emesa.60 Equally clear is the continuous tradition of the
cults of cities on the Phoenician coastline from the second millennium ..
into the Roman period. It is particularly significant that this was a conscious
survival. For in the first half of the second century .. Philon of Byblos
claimed to have composed his Phoenicica on the basis of a work in Phoenician (a language related to Aramaic and Hebrew) by one Sanchuniathon, who
dated from before the Trojan wars and who in fact perhaps belonged in the
Persian or early Hellenistic period and may have written in a dialect of Aramaic.61 Philon’s work is important both in showing that an educated Greek
could be explicitly conscious of the non-Hellenic traditions of his homeland and in filling out the very scanty documentary record of Phoenician
from this period. The record indeed hardly extends beyond the Hellenistic
age: we have for instance a Phoenician inscription from Oumm El-’Amed
dated to  ..,62 and another from Byblos which may be from the early Roman period, perhaps as late as the first century ..63 More securely dated, to
 .., is the latest of a series of bilingual Sidonian inscriptions from Athens
and the Peiraeus.64 About  .. Meleager, in an epigram recording his
birth at Gadara, his move to Tyre, and his old age in Cos, neatly contrasts the
form of greeting in Aramaic, Phoenician, and Greek.65 In the third century
. See A. Dupont-Sommer and L. Robert, La déesse de Hiérapolis-Castabala, Cilicie (),
relating a fourth-century .. Aramaic inscription to documents of the classical period.
. See H. Stocks, ‘‘Studien zu Lukians ‘de Syria dea,’ ’’ Berytus  (): ; G. Goossens,
Hiérapolis de Syrie: essai de monographie historique ().
. For the essentials, see RAC, s.v. ‘‘Elagabal.’’
. For the Phoenicica of Philon of Byblos, see Jacoby FGrH  F. –; on Sanchuniathon, see RE, s.v. ‘‘Sanchuniathon,’’ and M. L. West, Hesiod, Theogony (), –; cf., e.g.,
O. Eissfeldt, ‘‘Art und Aufbau der phönizischen Geschichte des Philo von Byblos,’’ Syria 
():  Kleine Schriften III (), . Note especially, on both the survival of Phoenician gods and the work of Philo, le Comte du Mesnil du Buisson, Études sur les dieux
phéniciens hérités par l’Empire romain ().
. M. Dunand and R. Duru, Oumm El-’Amed, une ville de l’époque hellénistique aux échelles
de Tyre (), no.  H. Donner and W. Röllig, Kanaanäische und aramäische Inschriften 2 I–III
(–), no. .
. Donner and Röllig (n. ), no. . Compare, however, J. Brian Peckham, The Development of the Late Phoenician Scripts (), .
. See Peckham (n. ), .
. Anth. Pal. , , lines –; A. S. F. Gow and D. L. Page, The Greek Anthology: Hellenistic Epigrams I (), : ‘‘whom heavenborn Tyre and Gadara’s holy soil reared to manhood
Paul of Samosata

we may note that Porphyry of Tyre gained this name, by which he is always
known, from a word-play by Longinus (whose mother came from Emesa)66
on his original name, ‘‘Malchus.’’ 67 Earlier, as Porphyry himself explains, his
friends had been accustomed to nickname him simply ‘‘Basileus.’’ 68 It is clear
at least that Porphyry and his friends knew that in his native tongue ( patrios
dialektos) MLK meant ‘‘King.’’ Whether he had a real knowledge of Phoenician (or Aramaic), and whether his studies of the Old Testament involved
any knowledge of Hebrew, remains obscure.69
Further north on the Phoenician coastline, from the cities around Aradus
and up to Gabala, we may note the presence of Semitic lettering on coins of
the Hellenistic period,70 the bilingual inscription of a man from Aradus who
died at Demetrias in Thessaly about  ..,71 and also a bilingual inscription (IGLS ) from Aradus itself, dating to / .. But thereafter there
is a gap of centuries before we learn from Socrates that Severianus, bishop of
Gabala in the early fifth century, though supposedly educated, spoke Greek
with difficulty and in a definite Syrian accent.72
The best confirmation, however, of the possibility that Philon of Byblos
might have known Phoenician or Aramaic comes from a remarkable source,
a scholion found in one of the manuscripts of Photius’ account of Iamblichus, the Greek novelist of the second century .. The information is represented as coming from Iamblichus himself and has certainly every appearance
of being circumstantial. According to the passage, Iamblichus recorded that
he was a Syrian on both his mother’s and his father’s side—‘‘not one of the
Greeks inhabiting Syria, but one of the natives, speaking their language and
living by their customs.’’ He acquired his knowledge of Babylonian lore from
and beloved Cos of Meropes tended his old age. If you are Syrian, Salam! If you are Phoenician, Naidios! If you are Greek, Chaire! And say the same yourself !’’ (Loeb translation).
. Suda .
. Eunapius, Vit. Soph. .
. Porph., Vit. Plot. : ‘‘ ‘Basileus’ is my [Porphyry’s] name; I am called Malkos in my
native tongue, for this name also my father used to call me. And ‘Basileus,’ the translation
of Malkos, if anyone wishes to change it into Greek.’’ Cf. Vit. Plot. , .
. Cf. J. Bidez, Vie de Porphyre (), –. I have not, however, found any serious discussion of this question. See F. Millar, ‘‘Porphyry: Ethnicity, Language and Alien Wisdom,’’
in J. Barnes and M. T. Griffin, eds., Philosophia Togata II: Plato and Aristotle at Rome (),
– ( chapter  of the present volume).
. H. Seyrig, ‘‘Monnaies hellénistiques,’’ Rev. Num.  (): , esp. –, –.
. O. Masson, ‘‘Recherches sur les Phéniciens dans le monde héllenistique,’’ BCH 
(): .
. Socrates, Hist. Eccles. , , .

Rome and the East
a captive taken in Trajan’s Parthian campaign; later he was trained in Greek
and became a skilled orator.73
This is the only item of evidence from a pagan source which clearly implies the existence of a whole class of educated Aramaic-speaking persons
in Roman Syria, and as such is of great importance. For further evidence
on the use of Aramaic in Syria we have to wait until the Christian period,
whose literature, as always, allows us insight into social levels which pagan
literature tended systematically to ignore.74 Though Christian writers call
the native language of Syria ‘‘Syrian,’’ it was actually what we call Aramaic,
and it will save confusion to reserve ‘‘Syriac’’ for the dialect of Edessa, the
script associated with it, and the literary language developed from it. In assembling the evidence it will not be necessary to pay special attention to
those passages where Christian writers merely refer generally to Aramaic or
to the meanings of individual words in it;75 what is important is to examine
those items which give some indication of the geographical or social range
of Aramaic-speaking, and its role within the church.
It so happens that much of the relevant evidence relates to the gentile
population of Palestine and its environs. The earliest Christian evidence
comes from the ‘‘long,’’ Syriac, recension of Eusebius’ Martyrs of Palestine,
and relates to Procopius, a martyr from Scythopolis executed in , who
had the role of interpreting into Aramaic in the church there.76 What is
meant by this is revealed by the detailed report of the conduct of services in
Aelia ( Jerusalem) made a century later by the pilgrim Egeria (or Aetheria).
She found that part of the congregation spoke only Greek, part only Aramaic, and part both. But the bishop, even if he knew Aramaic, spoke only
Greek in conducting services, while a presbyter had the task of translating
his words into Aramaic. Similarly, readings from the Bible were made first
in Greek, and then interpreted.77 Nothing could demonstrate more clearly
the values attached to the two languages. One may compare with it the
closely related evidence of Jerome, describing the funeral of Saint Paula in
. Photius, Bib. , (b), ed. Henry, II, .
. For collections of relevant passages in Christian writers, see C. Charon, ‘‘L’origine
ethnographique des Melkites,’’ Echos d’Orient  (): , ; G. Bardy, La question des langues
dans l’église ancienne (), –; A. H. M. Jones, The Later Roman Empire (), .
. See, e.g., Jerome, Vita Pauli  (PL XXIII, ); In Esaiam , , – (CCL LXXIII,
); Theodoret, Hist. Eccles. , , ; , , ; , , ; Hist. Relig.  (PG LXXXII, , ).
. For the text, see B. Violet, Texte und Untersuchungen XIV,  (),  and ; cf. H.
Delehaye, Les légendes hagiographiques (), ff.
. Pereg. Egeriae , – (CCL CLXXV, ); ed. H. Pétre, Éthérie, Journal de Voyage,
Sources Chrétiennes  (), –.
Paul of Samosata

Bethlehem in —‘‘the Psalms were sung in Greek, Latin, and Syriac in
this order.’’ 78
Towards the middle of the fourth century we find the hermit Hilarion, in
the desert outside Gaza, questioning a Frankish officer of Constantius in Aramaic, and miraculously causing the man to answer likewise; Hilarion then
repeated his question in Greek. He was clearly in fact bilingual; he came
from a village near Gaza and in his youth had been to study with a grammateus in Alexandria. Aramaic was evidently common, but not universal, in
the towns of this area. Retailing an incident when Hilarion visited Elusa,
Jerome emphasises that the place was semibarbarus: a large crowd abandoned
a festival of Venus which was in progress there and greeted Hilarion with
shouts of Barech—‘‘Bless [us].’’ 79 In  we find in Gaza itself a child, seized
by divine possession, calling out in Aramaic instructions for the burning of
the temple of Marnas there, and repeating them in Greek when interrogated
by the bishop, Porphyry. But, when interrogated herself, his mother swears
that neither she nor her son knows Greek.80
From Syria proper we do not have so much clear evidence. But Jerome
speaks of a hermit named Malchus living in a village thirty miles east of Antioch—‘‘A Syrian [Syrus], by nationality and language, as a native of the same
place would be,’’ though in fact he says later that the man was by origin an
immigrant from Nisibis.81 But what of Antioch itself ? Theodoret, who was
born and educated in Antioch, and was bishop of Cyrrhus in the first half
of the fifth century, wrote entirely in Greek but certainly knew Aramaic
and used his knowledge in interpreting the Bible.82 It is not likely, however,
that he learnt it in Antioch itself. For the clearest indication of the linguistic situation there comes from John Chrysostom, who in a sermon preached
in the city describes the country people coming in for a Christian festival as
a ‘‘people different from us so far as their language is concerned, but kin to
us in their faith.’’ 83 It seems indisputable that he means that, characteristi. Jerome, Ep. ,  (PL XXII,  CSEL LV, ).
. Jerome, Vit. Hil. , , ,  (PL XXIII , –).
. Marc. Diac., Vit. Porph. –, ed. H. Grégoire and M.-A. Kugener ().
. Jerome, Vita. Malchi – (PL XXIII, ).
. That he was a native Aramaic speaker would be an improper deduction from the
modesty of his own claim to Greek culture in Graec. affect. curatio  (PG LXXXIII, ). But
that he understood spoken Aramaic is clear from the incident in Hist. Relig.  (PG LXXII,
); cf. n. .
. John Chrys., Hom.  ad pop. Ant. I (PG XLIX, ); cf. Serm. de mart.  (PG L, ).
On the other hand, Hom. Matt. ,  (PG LVII, ), seems to imply the presence of some
Aramaic speakers in his audience, hence in the city.

Rome and the East
cally at least, the Christians of Antioch spoke Greek and those of its territory
(chora) Aramaic. Similarly, Theodoret mentions a hermit from the territory
of Cyrrhus who knew no Greek.84
Thus, though the evidence is slight and scattered, it is sufficient to show
that Aramaic was a living language in Roman Syria. But all the indications
are that it remained a rustic vernacular with no claim to rival Greek as a language of culture;85 it does not seem to have been until the fifth century that
Syriac came to be the vehicle of literature written in Roman Syria.86 Furthermore, although the appearance of a Christian Syriac literature in Edessa in
the second and third centuries is something of great interest and importance,
it seems likely that we should see it as an off-shoot of, rather than as a rival to,
Christian Greek culture. In short, just as it seems unlikely that either a man
from Samosata or a ruler of Palmyra could have seen himself as in any sense
representative of the ‘‘Orient’’ as against the Graeco-Roman world, so there
is very little to indicate that such a claim would have evoked any response in
Roman Syria. More particularly, the most we could claim from parallel and
later evidence for the church at Antioch is that in the third century it may
have begun to penetrate to the non-Hellenised strata of the population.
These considerations must tend to call in question certain presuppositions
from which the events of the s and s have been approached.That done,
it is time to consider the course of these events themselves.
Palmyra and Antioch
If we turn from the complex and elusive questions of the cultural background
to the more immediate and concrete political setting, the primary question
is chronological—when did Palmyrene control of Antioch begin?
The relevant events begin with the capture of Valerian in / by Shapur I, and the subsequent capture of Antioch.87 After this campaign Shapur
. Theod., Hist. Relig.  (PG LXXXII, , ).
. Note, however, Theodoret’s account of a fourth-century monastic foundation near
Zeugma where the original group of Greek-speaking monks was soon followed by one of
Aramaic-speakers, which was kept separate but had complete parity with the first, in Hist.
Relig.  (PG LXXXII, –).
. See Duval (n. ), ; Baumstark (n. ), ff.; I. Ortiz di Urbina, Patrologia Syriaca
(), chap. V.
. The date  rather than  is argued by S. Lopuszanski, La date de la capture de
Valérien et la chronologie des empereurs gaulois (); cf. Th. Pekáry, ‘‘Bemerkungen zur Chronologie des Jahrzehnts – n. Chr.,’’ Historia  (): , and PIR 2 L .
Paul of Samosata

retired, attacked en route by Odenathus of Palmyra.88 At the same time there
came the proclamation as Augusti of Macrianus and Quietus by their father,
Macrianus, and the praetorian prefect Ballista (Callistus).89 All that is necessary to note in the present context is that their regime lasted at least until 
(as revealed, for instance, in a papyrus dated to  May in their second joint
consulate and first Egyptian year, so ).90 Macrianus advanced into Europe
and was defeated by Gallienus. Quietus was defeated by Odenathus. It is important to note that our admittedly scanty sources say that the victory took
place at Emesa.91 Was the Palmyrene presence on the upper Orontes at this
moment followed by either a physical occupation or an effective overlordship of the cities of Roman Syria? Documentary evidence shows that by 
Odenathus had the title hypatikos (consular)—but this does not (as has been
claimed) serve to prove that he was governor of Syria-Phoenice.92 In about
 he is alleged (see below) to have received the title corrector totius Orientis (‘‘controller’’ or perhaps ‘‘restorer of the entire East’’) from Gallienus, and
almost certainly assumed rather than was given that of ‘‘king of kings.’’ The
available evidence on his movements and activities shows, however, nothing further concerned with the Roman province, but rather two invasions of
Mesopotamia, reaching to Ctesiphon.93 He is first heard of in Roman Syria
again at the moment of his murder, again at Emesa,94 which took place in
/.95 During this period it is notable that the mint of Antioch continues
. The evidence is late—Festus, Brev. ; Jerome, Chron., ed. Helm, ; HA, Trig. Tyr.
, –; Vit. Val. , –; Malalas, Chron. ,  Dindorf; Syncellus,  (Bonn); Zonaras ,
; J. Février, Essai sur l’histoire politique et économique de Palmyere (), –; J. Starcky,
Palmyre (), ff. The essential modern treatment of the chronology of Odenathus and
Vabalathus and their successive titulatures is D. Schlumberger, ‘‘L’inscription d’Hérodien:
remarques sur l’histoire des princes de Palmyre,’’ Bull. d’Ét. orient.  (–): .
. For the best account see A. Alföldi, ‘‘Die römische Münzprägung und die historischen Ereignisse im Osten zwischen  und  n. Chr.,’’ Berytus  ():  Studien zur
Geschichte der Weltkrise des . Jahrhunderts nach Christus (), .
. P. Oxy. XXXIV .
. Zon. , ; cf. HA, Vit. Gall. , , and Petrus Petricius, FHG IV,  Dio, ed.
Boissevain III, .
. Contra Schlumberger (n. ), , and J. Starcky (n. ), .The title appears in IGR III,
 J. Cantineau, Inventaire des inscriptions de Palmyre III (), no. . Cf. D. Magie, Roman
Rule in Asia Minor (), chap. XXIX, n. .
. See Février (n. ), ff.; Alföldi (n. ), ff. (ff.).
. Zosimus , , . Note, however, the variant tradition of Syncellus , – (Bonn),
according to which he was killed in Cappadocia on his way to repel a Gothic invasion.
. Probably between  August  and  August , because Alexandrian coins

Rome and the East
to produce coins for Gallienus from  up to the moment of his death in
, and for Claudius up to ; the coinage provides no suggestion of the
detachment of Syria from the Roman Empire up to this time.96 Confirmation
both of normal minting at Antioch up to  and of unsettled circumstances
immediately after that is provided by a hoard from near Bogazköy which
includes  coins of Gallienus, and  of Claudius minted at Antioch.97
The same story is told by both the literary sources and the documentary
evidence on titulature. Vabalathus began by assuming the titles, corrector (see
below) and ‘‘king of kings.’’ In  he advanced to consul, dux Romanorum,
and Imperator, though acknowledging Aurelian as Augustus. His own proclamation as ‘‘Augustus’’ came after August , when Zenobia appears without
the title ‘‘Augusta’’ on an inscription from Palmyra.98 The coinage of Antioch confirms that it did not come until the spring of , almost at the very
moment of Aurelian’s reconquest.99
The decisive break, however, and the Palmyrene occupation of Syria, had
clearly happened before this. Exactly how soon, it is not easy to say. Zosimus
portrays Aurelian, after the Danubian campaign of his first year (),100 reckoning with the Palmyrene occupation both of Egypt and of all Asia Minor up
to Ankara.101 Earlier, Zosimus places the invasion of Egypt under Claudius
(–) and says that Syrian troops were used for it;102 this squares with
the evidence of Malalas (, ) that Zenobia took Arabia from the Romans
seem to show the fourth year of Vabalathus ending on  August . See Schlumberger
(n. ), .
. See Alföldi (n. ), and CAH XII, –. Cf. C. Brenot and H.-G. Pflaum, ‘‘Les
émissions orientales de la fin du IIIe siècle après J.-C. à la lumière de deux trésors découverts
en Syrie,’’ Rev. Num.  (): ; cf. J.-P. Callu, La politique monétaire des empereurs romains
de  à , Bib. Ec. Fr. Ath. Rom.  (): –.
. K. Bittel, ‘‘Funde im östlichen Galatien: ein römischer Münzschatz von Devret,’’ Ist.
Mitt.  (): .
. J. Cantineau, Inventaire III (), no.  OGIS ; see Starcky (n. ), .
. See Schlumberger (n. ), and H. Seyrig, ‘‘Vabalathus Augustus,’’ Mélanges Michalowski (), . Note, however, an ostracon, O. Mich.  (Greek Ostraca in the University of
Michigan Collection III []; see Michigan Papyri VIII []), which dates to May–June 
and describes Aurelian and Athenodorus (Vabalathus) as Augusti.
. Dated to  by A. Alföldi, ‘‘Über die Juthungeneinfälle unter Aurelian,’’ BIAB 
():  Studien zur Geschichte der Weltkrise des . Jahrhunderts nach Christus (), .
But see n.  below.
. Zos. , , .
. Zos. , ; cf. HA, Claud. , –. Compare the inscription published by H. Seyrig,
Syria  (): –. Originating probably from the Hauran, it refers to the deaths of
many persons in Egypt, probably men recruited by Palmyra.
Paul of Samosata

under Claudius. The Egyptian evidence shows the invasion coming in the
regime of the prefect Probus in /;103 papyri and Alexandrian coins suggest that the actual capture of the city did not take place until the end of
.104 This agrees absolutely with the fact that the offerings the Palmyrenes
made at the shrine of Aphrodite at Aphaca, between Baalbek and Byblos,
were presented only a year before their final disaster in ,105 and that the
inscription of Zenobia from near Byblos calls her ‘‘Augusta,’’ and therefore
seems to belong in the latest period, probably .106
From this evidence it is not inconceivable that the Palmyrene drive southwest, through Arabia to Egypt, came before the occupation of northern
Syria. It is possible that Antioch was not occupied until ; none of our
evidence firmly indicates any earlier date.107 We must accept that Emesa was
firmly within the Palmyrene sphere of control through the s; but the territory of Emesa bordered that of Palmyra itself, and the two cities had close
cultural ties.108 Have we any evidence to suggest the existence of a wider
sphere of Palmyrene influence throughout this period?
Firstly, we have the title which is alleged to have been held by Odenathus from about , and inherited by Vabalathus in /, namely corrector
totius Orientis.109 If genuine, such an expression might indeed give some justification for the notion of an active Palmyrene patronage in Syrian cities.
But nothing in our documentary evidence justifies the supposition of such a
Latin title. The documents show that Odenathus had the title MTQNN’ DY
MDNH’ KLH.110 MTQNN’ could be the equivalent of Restitutor (restorer)
. See A. Stein, Die Präfekten von Agypten in der römischen Kaiserzeit (), –.
. See P. J. Parsons, ‘‘A Proclamation of Vaballathus?,’’ Chron. d’Ég.  (): . See
also n.  below.
. Zos. , .
. OGIS  IGR III . On the other hand, the milestone of Vabalathus on the
Bostra-Philadelphia road (AE , ) calls him ‘‘Im[perator] Caesar’’ but not ‘‘Augustus,’’
and will be a year or two earlier.
. Perhaps the nearest to concrete evidence available are the two tesserae (tokens) of
Herodianus and Zenobia which were probably found at Antioch, published by H. Seyrig,
‘‘Note sur Hérodien, prince de Palmyre,’’ Syria  (): . Herodianus should be the Herodes, son of Odenathus, who was killed with his father in / (HA, Trig. Tyr. –). But
even if the find-spots of the tesserae were certain, they are portable objects.
. See H. Seyrig, ‘‘Caractères de l’histoire d’Émèse,’’ Syria  (): .
. See M. Clermont-Ganneau, ‘‘Odeinat et Vaballat, rois de Palmyre, et leur titre
romain de Corrector,’’ RB  (): ; Février (n. ), ; A. Stein, Aegyptus  ():
; Schlumberger (n. ), , n. .
. Inventaire III  ( CIS ).

Rome and the East
just as well as of Corrector.111 The strict equivalent of Corrector appears only
in an inscription of Vabalathus,112 where he is called ’PNRTT’ DY MDYTH
KLH, where ’PNRTT’ is certainly a transcription of the Greek epanorthōtēs.
The two expressions are not necessarily identical in meaning. In any case
the expression which follows here—MDYTH—is the equivalent of polis or
‘‘provincia.’’ Even granted that KLH means ‘‘whole’’ or ‘‘every,’’ we have still
no indication of how wide was the geographical area concerned, still less
any justification for assuming an equivalence and translating Odenathus’ title
as corrector totius Orientis. On the contrary, Zosimus, in his detailed and circumstantial narrative, has an anecdote of the Palmyrenes, evidently during
their apogee in –, enquiring of Apollo at Seleucia ‘‘whether they would
obtain the domination of the East.’’ 113
The literary sources provide no firmer basis for supposing any established
Palmyrene hegemony over any of the eastern Roman provinces. Zosimus and
Zonaras state no more than that Gallienus gave Odenathus a major military
role against Persia.114 Eutropius and Orosius, for what they are worth on a
precise point, imply that Zenobia’s wider ambitions post-dated the murder
of her husband.115 In fact, the notion of a general rule of the East by Odenathus depends fundamentally on a number of grandiose generalizations in
the Historia Augusta 116—which also states, falsely, that Gallienus awarded him
the title ‘‘Augustus.’’ 117
That Palmyra or its rulers exercised any real influence in Antioch before
about  thus remains a pure speculation unsupported by any reliable concrete evidence. On the other hand, we do have the testimony of Zosimus to
the fact that when Aurelian retook Antioch in  there was a pro-Palmyrene
group there which was preparing to flee in terror until the Emperor issued an
. See J. Cantineau, ‘‘Un Restitutor Orientis dans les inscriptions de Palmyre,’’ Journal
Asiatique  (): .
. See Clermont-Ganneau (n. ), , .
. Zos. , , .
. Zos. , , ; Zon. , . Cf. Syncellus,  (Bonn.): ‘‘He was appointed also commander of the East by Gallienus because of this.’’
. Eutropius , , : ‘‘Zenobia ruled the East after the murder of her husband Odenathus’’; Orosius , , : ‘‘Zenobia, who after the murder of her husband claimed possession
of Syria for herself ’’; cf. Festus, Brev. : ‘‘For after her husband’s death she held imperium
over the East under the command of a female.’’
. HA, Vit. Gall. , : ‘‘When Odenathus already obtained imperium over the East’’; ,
: ‘‘Odenathus became emperor of nearly the whole East’’; , : ‘‘Odenathus king of the
Palmyrenes obtained imperium over the whole East’’; Trig. Tyr. , : ‘‘Odenatus, who once
already held the East.’’
. HA, Vit. Gall. , .
Paul of Samosata

edict of amnesty;118 and, on the other side, Jerome records the name of an apparently Antiochene commander who fought at the battle of Immae against
Zenobia.119 These statements certainly make it reasonable to ask whether the
career of a controversial figure in Antioch at this time can be explained in
terms of divided political loyalties. Whether this was so in the case of Paul
must depend on a detailed examination of the evidence about him.
Paul in Antioch
By far our most reliable and valuable evidence comes from book  of Eusebius’ Ecclesiastical History, which gives substantial excerpts from the letter sent
by the synod which deposed Paul, addressed to Dionysius, bishop of Rome,
and Maximus, bishop of Alexandria. It is noteworthy that both in his own
remarks and in his choice of excerpts Eusebius concentrates (conveniently
for our purposes) on the externals of Paul’s conduct, and is almost entirely
silent on the precise heresies as to the nature of Christ of which he was held
guilty. For Paul’s views—he affirmed the unity of God and the Word (using
the term homoousios) and denied the divinity of Christ—we have to rely on
late reports and quotations of the dialogue between himself and Malchion
(see below) at his examination.120 Eusebius’ delicacy on this point can hardly
be unrelated to the fact of his own involvement with Arianism;121 that the
historical connection between Paul’s doctrines and Arianism has sometimes
been denied by modern scholars is less relevant than the fact that it was vigorously asserted by a contemporary, Peter, bishop of Alexandria.122
The chronological framework is crucial to our understanding of the career. Firstly, Eusebius (HE , , ) records the election of Dionysius to succeed the martyred Xystus as bishop of Rome: Xystus was executed on  August ,123 and Dionysius was consecrated on  July  or .124 In the
. Zos. , , .
. Jer., Chron., ed. Helm, : ‘‘In this battle Pompeianus, nicknamed Francus, held
command against her. His family survives to this day in Antioch. Eugarius, a priest, most
dear to us, is a descendant of his.’’
. See H. de Riedmatten, Les Actes du procès de Paul de Samosate; étude sur la Christologie
du III e au IV e siècle ().
. For a survey, D. S. Wallace-Hadrill, Eusebius of Caesarea (), chap. VI.
. See the letter of Alexander to his namesake, bishop of Constantinople, Theodoret,
HE , , –.
. Cypr., Ep. , . Cf. L. Duchesne, Le Liber Pontificalis I (), .
. See Duchesne (n. ), ; C. H. Turner, ‘‘The Papal Chronology of the Third
Century,’’ J. Th. St.  (–): ; E. Caspar, Geschichte des Papsttums I (), –.

Rome and the East
next sentence Eusebius gives the death of the bishop Demetrianus at Antioch, and his succession by Paul; a later legend, too readily believed, relates
that Demetrianus was in fact carried into captivity by Shapur.125 Jerome’s
Chronicle places the election of Paul in .126 Paul’s heretical tendencies were
evidently not long in showing themselves, for a synod convened to consider
them some three or four years later. The approximate date can be confidently
established: among the bishops invited was Dionysius of Alexandria, who
excused himself on the grounds of age and feebleness, and wrote a letter exposing Paul’s errors, but then ‘‘at that time [he] died, in the twelfth year of
Gallienus’’ (/ or /).127 There is no reason to doubt that the synod
took place in about . The upshot of the synod was a promise by Paul, not
fulfilled, to abandon his errors.128
The date of the decisive synod is more difficult to establish. In HE , , ,
Eusebius records together the death of Gallienus (), the reign of Claudius,
and the accession of Aurelian (). He then passes on, with the expression
‘‘at that [i.e., time],’’ to the affair of Paul (, , ). Various converging items
of evidence combine, however, to suggest that the synod in fact took place
over the winter of /.129 Jerome’s Chronicle 130 places the deposition of Paul
in the year before the nd Olympiad (), and this date is reflected also
in Zonaras, who makes the episode approximately contemporary with the
death of Gallienus.131 In the declaration against Nestorius posted up at Constantinople in /, Eusebius, the later bishop of Dorylaeum, referred to the
excommunication of Paul  years before 132—so, if taken precisely, /.
The letter of the synod itself contains two clues. It is addressed to Dionysius, bishop of Rome, who died on  or  December .133 It may be that
. The tale is given in the Arabic Chronicle of Seert (Patr. Or. IV, –). See Bardy
(n. ), ff., and Downey (n. ), , and M. L. Chaumont, ‘‘Les Sassanides et la christianisation de l’Empire iranien au IIIe siècle de notre ère,’’ Rev. Hist. Rel.  (): . There
is not the slightest reason to prefer such a source to the plain statement of Eusebius.
. Jerome, Chron., ed. Helm, .
. Euseb., HE , , ; ,  (cf. , ). For the problem of Gallienus’ regnal years, see
E. Manni, ‘‘Note di epigrafia gallieniana,’’ Epigraphica  (): .
. Euseb., HE , , .
. See Loofs (n. ), ff.; Bardy (n. ), –.
. Ed. Helm, .
. Zon. , .
. See de Riedmatten (n. ), , . For the text, see E. Schwartz, Acta Conciliorum I.., para.  (p. ).
. See O. Bardenwehr, Geschichte der altkirchlichen Literatur II (), –; cf. C. H.
Turner (n. ), –; L. Duchesne (n. ), .
Paul of Samosata

the letter of the synod arrived after his death; for though the surviving letter
of his successor, Felix, addressed to Maximus of Alexandria on precisely the
point at issue, the divinity of Christ, is generally regarded as an Apollinarist
forgery, it may have replaced a genuine original letter.134 Secondly, the letter
refers (, ) to the fact that Firmilian, bishop of Caesarea in Cappadocia,
died at Tarsos on his way to Antioch, at a moment when the other participants had arrived and were waiting for him; according to the Greek calendar
his death took place on  October.135 The date of the synod can therefore
be fixed with considerable precision and with no real room for doubt.
The central element in the view that connects the career of Paul with
the wider political history of his time is a passage in Eusebius’ quotation
of the synodal letter.136 The bishops state that Paul came from an impoverished family but had made himself rich through extortion in his bishopric—‘‘by means of lawless deeds and sacrilegious plundering and extortions
from brethren by threats . . . making the service of God [theosebeia] a source
of profit.’’ Then comes the key phrase—‘‘he sets his mind on high things
and is lifted up, clothing himself in worldly goods and wishing to be called
ducenarius [see below] rather than bishop.’’ The letter continues by describing the public role to which these aspirations led him—parading across the
market-places, reading letters and answering them as he went, with a numerous bodyguard, some marching in front and some behind, so that his pomp
and pride made the faith an object of envy and hatred; in meetings of the
church he devised various means to impress the simple-minded—‘‘arranging for himself a tribunal [bēma] and a high throne . . . and having a sēkrēton
[audience chamber; see text to n.  below] and calling it that, like worldly
rulers.’’
The term ducenarius was a well-established expression, deriving from the
level of salary, for a high-ranking imperial procurator.137 We find it used in a
letter of Cyprian and his fellow bishops to the congregation of Emerita—
‘‘proceedings taking place in public before the procurator ducenarius.’’ 138 As used
in Antioch, it could in normal times only have referred to the procurator of
. See, e.g., A. von Harnack, Geschichte der altchristlichen Literatur bis Eusebius I (),
ff.; cf. J. Quasten, Patrology II (), .
. See Loofs (n. ), ; Bardy (n. ), .
. HE , , –.
. See H.-G. Pflaum, Les procurateurs équestres (), ff.; and Les carrières procuratoriennes (–), –, with F. Millar’s review in JRS  (): – ( ‘‘The Equestrian Career under the Empire,’’ chapter  in Fergus Millar, Rome, the Greek World, and the
East II: Government, Society, and Culture in the Roman Empire).
. Cypr., Ep. , .

Rome and the East
the province of Syria Coele.139 Like the parallel term centenarius (a procurator
with a salary of , sesterces per annum),140 however, it is also attested in
use at Palmyra, being one of the many titles of a prominent Palmyrene, Septimius Vorodes.141 Moreover, but for a crucial textual difficulty, there would be
confirmation from Cyprian for the notion that bishops might combine with
their office the holding of an imperial procuratorship. In his On Those Who
Lapsed, , Cyprian writes: ‘‘The bishops . . . spurning a divine procuratorship, become procurators of worldly kings [so SW; of worldly affairs, R],
neglecting their episcopal thrones and deserting the people.’’
The logic of the passage seems to demand the reading ‘‘affairs’’; there is
nothing else in it which refers in any way to the emperors, and the other
activities of the delinquent bishops seem to represent entirely private profiteering. This evidence can therefore not be adduced in support.
Moreover, whatever interpretation is given to the crucial passage of the
synodal letter, one thing is clear—namely, that it contains no reference to
Palmyra or its rulers. This may now seem hardly surprising for, as we have
seen (text to nn. – above), it was written before the period from which
we have any evidence of a Palmyrene military presence in Antioch, and in
a period when no other evidence proves a general Palmyrene patronage of
the Syrian cities. None the less, the connection with Palmyra, missing in the
letter, is duly supplied by later Christian sources. Tabulated in chronological
order, they are as follows:142
. Athanasius, Historia Arianorum , : ‘‘Zenobia was a Jew and she
patronised Paul of Samosata’’ (c. ). (Cf. also the fragment, possibly of
Athanasius, in PG XXVI, .)
. Filastrius, Diversarum haereseon liber, / (CSEL XXXVIII, ;
CCL IX, ): ‘‘He preached that Christ was a just man, not a real
God . . . hence he himself also taught Zenobia, at that time a queen
[regina] in the East, to become a Jew’’ (–).
. John Chrysostom, Hom.  in Joannem (PG LIX, col. ): ‘‘For not
unwittingly, but knowing full well, he transgressed, suffering the same
thing as the Jews. For like the latter, who looking to men gave up the
soundness of the faith, although on the one hand, knowing that he is
the sole begotten son of God, on the other hand, not acknowledging
. Pflaum, Carrières .
. See Schlumberger (n. ), –.
. OGIS  IGR III, ; OGIS  IGR III, ; IGR III, ; cf. H. Ingholt,
‘‘Inscriptions and Sculptures from Palmyra,’’ Berytus  (): –.
. Cf. Loofs (n. ), ; Bardy (n. ), ff.
Paul of Samosata

this because of their rulers, lest they be expelled from the synagogues.
In the same way, they say, also this man [i.e., Paul] wishing to ingratiate
himself with a certain woman, gave up his own salvation’’ ().
. Theodoret, Haereticarum fabularum compendium ,  (PG LXXXIII,
col. ): ‘‘While Zenobia was ruling that region (for the Persians
having defeated the Romans, handed to her the rule over Syria and
Phoenicia), he [Paul] drifted into the heresy of Artemon, believing that
in this way he would flatter her who believed in what the Jews do’’
(c. ).
All of these passages bring in, in one form or another, a tendentious reference to Judaism. This is not surprising, for there was a clear resemblance
between Jewish belief and Paul’s teaching—and Epiphanius in the Panarion
(while making no reference to Zenobia) says that the Paulianistae differ from
the Jews only in not observing the Sabbath or circumcision.143 On the other
hand there is a separate account, apparently not related to the tradition about
Paul, of the Judaising tendencies of Zenobia. This appears in some remarks
made about Longinus by Photius 144—‘‘she also converted from her Greek
superstition to the customs of the Jews, as an old source writes.’’ Even of this
there is very little confirmatory evidence: the Talmud has one anecdote of
an appeal by Jewish elders to Zenobia, but the attitude expressed there is
otherwise hostile.145 On the other hand an inscription in Latin from Egypt
does show a regina and rex, evidently Zenobia and Vabalathus, confirming a
Ptolemaic grant of right of asylum to a synagogue.146
The most extreme version of the story, though the earliest attested, that
Zenobia was herself Jewish can be firmly discounted. It is true that we find
an indubitably Jewish Zenobius on a Palmyrene inscription of .. .147
But the name is common in Palmyra, and a more probable candidate for
relationship to Zenobia would be the Iulius Aurelius Zenobius whom we
find exercising important functions during the visit of Severus Alexander
in .148 Jewish sources show no awareness that Zenobia was Jewish, and
. Epiphanius, Panarion, , , .
. Photius, Bib. , ed. Bekker,  (see n.  above).
. Jerusalem Talmud, Terumoth : (trans. Schwab., III, ). See J. Neusner, A History
of the Jews in Babylonia II: The Early Sasanian Period (), .
. OGIS  ILS  Corp. Ins. Jud.  E. Gabba, Iscrizioni greche e latine per lo
studio della bibbia (), no. .
. Frey, Corp. Ins. Jud. II, no. .
. OGIS  IGR III ; cf. PIR 2 I, . It is not a fatal objection to this possibility
that Zenobia is found with the nomen ‘‘Septimia,’’ for this is not attested for Palmyrenes,

Rome and the East
the possibility is of course incompatible with the much better attested claim
that she had leanings towards Judaism. This in its turn, though it has little
positive support (see above), is not impossible. There is fairly substantial evidence for a Jewish community in Palmyra.149 Her possible favour to Judaism
combined with the nature of Paul’s doctrines may explain how the story of
their connection arose. But we can also discount the version of Theodoret
and Chrysostom, that the desire to please Zenobia was the cause of Paul’s
lapse into heresy. Paul was already accused of heresy in about , Zenobia
only became prominent (so far as we know) after the death of her husband in
/, and the Palmyrenes perhaps entered Antioch only in . At the most
then, her patronage of Paul may have begun in the period after the synod,
when Paul clung on obstinately in the church house.150 We cannot actually
disprove the third version of the story, that of Filastrius, that Paul influenced
Zenobia in the direction of Judaism.
What then of Paul’s procuratorship? A closer look will show it to be a
fantasy. The whole sense of the passage in the synodal letter is that Paul as
bishop modelled his style and public appearances after those of imperial officials—‘‘wishing to be called ducenarius rather than bishop,’’ ‘‘arranging for himself a
tribunal and high throne’’ (probably modelled on a governor’s tribunal and
seat, though even normal bishops had something of the sort),151 ‘‘having a
secretum, and calling it that, like worldly rulers’’ (the reference must be to what
is normally called the secretarium of a governor, the audience chamber where
a number of attested trials of martyrs took place).152 Everything that is said
of the improper activities of Paul relates to the life of the Christian congregation—extorting money from the brethren, making the service of God a
source of profit, organizing spectacles in the assemblies of the church, insulting
other than the family of Odenathus, until the s; see Schlumberger, Bull. d’Ét. Or. 
(–): .
. See E. Peterson, ΕΙΣ ΘΕΟΣ (), –; Frey, Corp. Ins. Jud. II, pp. –.
. So, in effect, Loofs (n. ), .
. Both thronos and bēma are attested for both civil and ecclesiastical authorities; see
E. Stommel, ‘‘Bischofsstuhl und Hoher Thron,’’ Jahrb. f. Ant. u. Chr.  (): . And note
that very similar podia are found at Dura in the Christian building and in the palace of the
Dux; see C. H. Kraeling, The Excavations at Dura-Europos, Final Report VIII, : The Christian
Building (), –. But note L. Robert, Hellenica IV (), –, on the theme of the
governor’s thronos and bēma in Greek epigrams.
. Note the trial of the Scillitan martyrs, ‘‘in the secretarium, in Carthage’’; of Crispina
‘‘in the colony of Thebestina, in the secretarium, when Anulinus the proconsul was (holding
court) at the tribunal’’; of Cyprian, ‘‘Carthagine in secretario,’’ R. Knopf, G. Krüger, and
G. Ruhbach, Ausgewählte Martyrakten 4 (), , , . Cf. RE, s.v. ‘‘Secretarium.’’
Paul of Samosata

those who received his words without excessive enthusiasm as was fitting in a
house of God. The worthy bishops would have been surprised to know how
long and undeserved a life their words have given to ‘‘Paul of Samosata, the
ducenarius of Zenobia.’’
The Appeal to Aurelian
The political history of his time will therefore explain neither the rise of Paul
of Samosata nor his formal deposition. Whether it helps to explain his refusal
to leave the church house and his opponents’ successful appeal to Aurelian,
it is more difficult to say. At first sight it seems an obvious supposition that
he clung on until Aurelian entered Antioch in , whereupon the loyalist
party petitioned the Emperor and had him ejected; as was mentioned above
(text to n. ), we even know that there were then Palmyrene supporters in
the city, whose fears Aurelian had to still, and also loyalists who aided the
Roman forces.
But it must be noted that there are three separate problems here. The first
concerns the motivation of the appeal to the Emperor; almost all that has
been written about this episode has assumed almost unconsciously that only
exceptional circumstances—or, more precisely, an immediate political situation such as that imagined here—would serve to explain how such an appeal could have been made. Secondly, there is the question of chronology;
how and when did the appeal reach Aurelian, who assumed power probably in the summer of  and did not reach Antioch until ? Thirdly,
there is the favourable response by Aurelian, and the remarkable terms in
which it was expressed. The first thing is to set out the brief couple of sentences of Eusebius (, , ) which are our sole evidence for these proceedings. After concluding his extracts from the synodal letter, and mentioning
the deposition of Paul and the election in his place of Domnus, Eusebius
continues:
But as Paul refused on any account to give up possession of the house
of the church, the emperor Aurelian on being petitioned, gave an extremely just decision regarding the matter, ordering the assignment of
the house to those with whom the bishops of the doctrine in Italy and
Rome would communicate in writing. Thus, then, was the aforesaid
man driven with the utmost indignity from the church by the ruler of
the world. (based on the Loeb translation)
Even the most immediate character of the situation is not easy to grasp. What
was ‘‘the house of the church’’? We can suppose that it bore some resem-

Rome and the East
blance to the private house at Dura-Europos, converted for use in Christian
services.153 Moreover, there is no doubt that by the end of the third century
Christian communities generally possessed a regular meeting place, variously
called oikos ekklēsias, kyriakon, or proseuktērion. There does not appear, however, to be good evidence from this period that it was ever combined with an
actual episcopal residence.154 One must suppose rather that Paul continued
to perform services there, perhaps with the support of a part of the congregation.
Hence the appeal to the Emperor. It is crucial to our understanding of
both the Roman Empire, and of the place of the church within it, to realize that we do not have to find exceptional political circumstances to explain
recourse to the emperor as arbiter. For an immense mass of evidence shows
us that individuals and communities saw the giving of justice as a primary
function of the emperor, just as they had of the Hellenistic monarchs who
preceded them.155 Even if we confine ourselves to Syrian evidence alone,
we may note, firstly, the inscription which contains a series of appeals to
rulers from the temple community of Baetocaece, stretching from probably
the third century .. to .. /.156 Then we have the words of an orator
addressing Caracalla in Antioch in  May , on the subject of the priesthood of the temple at Dmeir: ‘‘There is a famous temple of Zeus among
them, famous indeed among all the people of the area . . . they frequent it
and conduct processions to it. The first wrong done by our adversary . . . he
benefits from freedom from [taxation and?] liturgies, wears a gold crown,
[enjoys precedence?], wields a sceptre and has proclaimed himself priest of
. See C. H. Kraeling (n. ), esp. ff., ‘‘The Christian Building at Dura, and Early
Church Architecture’’; cf. R. Krautheimer, Early Christian and Byzantine Architecture (),
chap. I. Compare R. L. P. Milburn, ‘‘Ο ΤΗΣ ΕΚΚΛΗΣΙΑΣ ΟΙΚΟΣ,’’ J. Th. St.  ():
, and Chr. Mohrmann, ‘‘Les dénominations de l’église en tant qu’édifice en grec et en
latin au cours des premiers siècles chrétiens,’’ Rev. Sci. Rel.  (): supp. , esp. –.
. Though, for instance, Lebreton and Zeiller (n. ), , use the term ‘‘maison épiscopale.’’ Kraeling (n. ),  and , does not seem to me to offer concrete evidence for the
residence of presbyters in domus ecclesiae.
. See F. Millar, ‘‘The Emperor, the Senate, and the Provinces,’’ JRS  (): 
( chapter  of F. Millar, Rome, the Greek World, and the East I: The Roman Republic and the
Augustan Revolution; ‘‘Emperors at Work,’’ JRS  ():  ( chapter  of F. Millar, Rome,
the Greek World, and the East II: Government, Society, and Culture in the Roman Empire; The
Roman Empire and Its Neighbours (), –, –.
. OGIS  Abbott and Johnson, Municipal Administration, no. ; cf. F. Millar,
CR, n.s.,  (): , which requires correction in the light of H. Seyrig, ‘‘Aradus et
Baetocaecé,’’ Syria  (): . The document is now re-edited as IGLS IV, .
Paul of Samosata

Zeus. How he gained such a privilege I shall show.’’ 157 The close resemblance
to what will have been said to Aurelian about Paul of Samosata needs no
stressing. It does not follow of itself that a Christian community could have
come easily to make such a claim before a pagan emperor. But the groundwork for such an advance had in fact long been prepared. For just about a
century before the case of Paul arose, we find Theophilus, bishop of Antioch, discussing the Christian view of the emperor.158 Why do the Christians
not worship the emperor? Because he is not a god, but a man, appointed by
God, not to receive homage, but to give judgment rightly.
It must be remembered that Gallienus’ edict of toleration was by now
about a decade old. There was on the face of it nothing to stop the increasingly settled and well-established church from taking its place with other
institutions in the life of the Empire. The moral had indeed been drawn immediately after the edict. For Eusebius reproduces a rescript (antigraphē ) of
Gallienus, written in answer to a request from the bishops of Egypt for recovery of church property there; in the same paragraph he mentions what
was evidently another rescript to a different group of local bishops concerning the recovery of Christian cemeteries.159
It was therefore only a relatively small further step that the congregation
of Antioch should address themselves to the Emperor for help in recovering
their church house from an impudent heretic. To do so they did not necessarily have to wait until the Emperor was in the locality. We have no reason to
think that Gallienus ever visited Egypt, and the likelihood is that the Eygptian churches—like innumerable associations or cities before them—sent a
delegation to him.
What then was the sequence of events leading to the deposition of Paul?
Nothing in the admittedly extremely brief narrative of Eusebius prepares us
for the possibility that a period of three and a half years passed, with Paul
in illegal possession of the church house until Aurelian arrived in person in
Antioch. Moreover, Eusebius mentions the succession of Timaeus to Domnus, Paul’s successor, as bishop of Antioch, quite separately from the affair of
Paul—and Jerome puts Timaeus’ consecration precisely in , and before
the reconquest by Aurelian, which he puts in .160 So it is equally possible
. P. Roussel and F. de Visscher, ‘‘Les inscriptions du temple de Dmeir,’’ Syria  (–
): ; W. Kunkel ‘‘Der Prozess der Gohariener vor Caracalla,’’ in Festschrift H. Lewald
(), ; SEG XVII . On the lines in question (–), see N. Lewis, ‘‘Cognitio Caracallae de Goharienis: Two Textual Restorations,’’ TAPhA  (): .
. Theophilus, Ad Autolycum , II.
. HE , .
. HE , , ; Jerome, Chron., ed. Helm, .

Rome and the East
that after the conclusion of the synod—whose letter was written, at the latest,
before news of the death of Dionysius of Rome in December  had arrived,
and possibly as early as November of that year (see text to n.  above)—the
Antiochene congregation responded more urgently to Paul’s obduracy, and,
perhaps some time in , despatched a delegation to the Emperor.
The evidence of the early Constantinian period suggests that it normally
took several months from the issuing of an imperial reply to its arrival in
a provincial city.161 Similarly, though here our information is less precise, it
will have taken a comparable period for a delegation to reach the emperor
wherever he happened to be, to gain an audience, and to receive an answer.162
These were, it hardly needs to be said, troubled times. Claudius was mainly
engaged in combating barbarian invasions in central and eastern Europe up
to the moment of his death at Sirmium, which certainly took place later than
 December ,163 and seems now not to have been until summer, .164 It
would not be in the least surprising if, for one reason or another, no imperial
reply was forthcoming until given by Aurelian; the occasion could have been
any moment from the very beginning of his reign onwards.
. Cod. Just. , , ; , , ; , , ; , , ; , , . For reasons which remain
obscure, all these rescripts, to which the dates of propositio (posting up) at the comitatus (the
emperor’s court) and acceptance by the city concerned are attached, date from the decade
–.
. Though where our sources give some indication of the time spent on embassies,
it tends to be in exceptional cases. But note, e.g., Jos., Ant. , –—deliberate delay
by Tiberius; Philo. Leg. /—Jewish Alexandrian embassy waiting for an audience with
Gaius; Pliny, Pan. , –.
. Clandius’ third trib[unicia] pot[estas] is clearly attested, CIL II ; III  ILS
. See L. Bivona, ‘‘Per la cronologia di Aureliano,’’ Epigraphica  (): ; as Bivona
points out (), this essential datum is missing from J. Lafaurie, ‘‘La chronologie impériale
de  à ,’’ Bull. Soc. Nat. Ant. France (): ; for the evidence on Claudius’ activities,
see P. Damerau, Kaiser Claudius II Goticus, Klio, Beih. XXXIII, n.f.,  ().
. See J. Rea, ‘‘The Chronology’’ (of the Corn Dole Archive, P. Oxy. –) published in P. Oxy. XL (), pp. –, which the author was kind enough to show me
before publication. The scheme which emerges is as follows: () Claudius survived until
shortly before the end of his second Egyptian regnal year (/); () Quintillus is attested on Alexandrian coins but no known papyri are dated by him; () the third year of
Claudius is attested on coins and papyri, suggesting that the news of his death had not
yet spread—/ is therefore Quintillus I (coins), Claudius , and also (from December)
Aurelian I/Vabalathus ; () subsequently Aurelian re-numbered his Egyptian regnal years
to date from the death of Claudius, so making / Aurelian I; () Egypt seems to have
been recovered in the summer of .
Paul of Samosata

Even if, therefore, we suppose that the reply was not given in the context
of the reconquest of , would it not still have been influenced by political considerations? It is certainly very difficult, but not absolutely impossible
(text to n.  above), to argue that it could have been given by Aurelian in
ignorance of the Palmyrene occupation of Syria. None the less, it is notable
that the indication given by Eusebius of its terms contains no reference to
Palmyra. It is possible to interpret the description of the orthodox party—
‘‘those with whom the bishops of Rome and Italy were in communication’’—
as a veiled reference to their loyalty to Rome. But it cannot be emphasised
too strongly that, as it stands, the definition is strictly ecclesiastical. How then
did Aurelian arrive at this remarkable formulation? Two hypotheses, not incompatible, are possible. The first is that it is quite evident from a number
of examples of imperial letters that emperors in formulating a response very
often took the passive course of following closely the wording of the request
presented to them.165 Thus Aurelian may well have taken over a description
which the orthodox party gave to themselves. The second is that, whether or
not the delegation was finally heard in Italy, or Rome itself, it may have been
actively supported by bishops from there. For what it is worth, Zosimus, our
only more or less coherent narrative source, shows Aurelian setting out from
Rome at the beginning of his reign, going to Aquileia, then to Pannonia, and
subsequently returning to Italy.166 The delegation from Antioch could well
have obtained a hearing somewhere in Italy before or after these campaigns.
All this, however, remains a hypothesis. Aurelian’s decision may well not
have come until , and in either case the formulation of it may have related
to divisions in the Antiochene church, which themselves reflected political
allegiance to Rome or to Palmyra. All that can be asserted is that, if we set
what we are actually told by someone relatively close to the event against
the wider background of what we know of the nature of the Roman Empire, then we do not need an explanation in terms of contemporary politics
for either the appeal to Aurelian or his reply. The relevance of these considerations to the early contacts of the church and Constantine need not be
stressed.
. The parallel case of the repetition by the emperor of the wording of the original
letter when replying to a provincial governor is patent in the case of Pliny and Trajan; see
A. N. Sherwin-White, The Letters of Pliny (), ff. For letters in reply to city embassies,
see F. Millar, The Roman Empire and Its Neighbours (n. ), .
. Zos. , –. See Alföldi (n. ).

Rome and the East
Paul’s Heresy and Local Culture?
As so often, the effect of examining and dismissing large-scale assumptions
which have been too hastily accepted is only to replace them with more precise doubts and questions. For all that is argued above, can we really be sure
that the story of Paul does not reveal the suppression of a strain of local belief
and liturgical practice by the prevailing orthodoxy of the Greek church? We
must note, for instance, that among his opponents were men of established
reputation in contemporary pagan Greek culture. His principal Antiochene
opponent, the presbyter Malchion, was a learned man who was (apparently)
the chief teacher of rhetoric at Antioch;167 while among those who came
to Antioch to examine his case was Anatolius from Alexandria, a student of
arithmetic, geometry, astronomy, physics, and rhetoric, who was successively
head of the Aristotelian school at Alexandria and bishop of Laodicea.168
On the other side, the resemblance between Paul’s view of Christ and Jewish belief was (as noted above, text to nn. –) unmistakable. We cannot
prove that Paul was not influenced by members of the substantial Jewish community in Antioch;169 the example of Origen at Caesarea shows that learned
discussions between Jews and Christians were still possible.170 It may be noted
that two Antiochene Christians of the generation after Paul, the presbyter
Dorotheus 171 and the important theologian and biblical scholar, Lucian of
Antioch,172 are both attested as having had a profound knowledge of Hebrew. But whatever was said in later Christian literature about the Judaising
tendencies of Paul or his followers, this particular line of attack was not used,
so far as we know, by his contemporaries. On the contrary, they clearly regarded his heresy as a revival of that of Artemon (or Artemas);173 of the latter
. If that is what is meant by the puzzling phrase of Ensebius, HE , , , ‘‘who stood
at the head of a school of rhetoric, one of the Greek educational establishments at Antioch’’ (Loeb translation). Note M. Richard, ‘‘Malchion et Paul de Samosate: le témoignage
d’Eusèbe de Césarée,’’ Eph. Theol. Lovanienses  (): .
. HE , , , .
. See C. H. Kraeling, ‘‘The Jewish Community at Antioch,’’ Journ. Bib. Lit.  ():
.
. See, e.g., M. Simon, Verus Israel 2 (), ; H. Chadwick, Origen: Contra Celsum
(), .
. Euseb., HE , , –.
. Suda, s.v. Loukianos. See G. Bardy, Recherches sur Saint Lucien d’Antioche et son école
(), ff., ff.
. HE , , – (two extracts from the synodal letter); cf. Pamphilus, Apologia pro
Origene  (PG XVII, –); Theodoret, Haer. Fab. comp. II,  (PG LXXXIII, ).
Paul of Samosata

nothing is known, though a phrase in the synodal letter appears to imply that
he was still alive. What is clear is that the Adoptianist heresy here referred
to is (rather obscurely) described by Eusebius as having originated in Rome
in the late second century under the impulse of one Theodotus of Byzantium.174 If there is a ‘‘local’’ element in the nature of Paul’s heresy, it is rather,
perhaps, to be found in its resemblance to that of Beryllus of Bostra, who
thought that Christ did not pre-exist his birth and had no divinity except
that of the Father dwelling in him—and was duly corrected by an assembly
of bishops, assisted by Origen, in about –.175
Some claims for a local origin can be advanced for some of the innovations of Paul in liturgy and church practise. The synodal letter speaks of ‘‘the
females introduced into his house, as the Antiochians call them’’;176 the reference is to virgines subintroductae, that is (in theory), women living with priests
without sexual relations. It is sometimes suggested that this was a distinctively Syrian form of asceticism.177 But the practice, with its associated scandals, is clearly attested a few years earlier in Africa.178 Paul also had a chorus
of women who sang psalms specially composed in honour of himself—and
were alleged to proclaim that he was in fact an angel who had descended from
heaven.179 Here one can only note, for lack of detail in the account of Paul,
that other evidence indicates a particularly rich tradition of hymn composition in Syriac, beginning with Bardesanes and his son Harmonius (see text
to nn. – above), and the hymns incorporated in the Acts of Thomas.180 It
is at least a reasonable speculation that Paul’s compositions were related to
this tradition.
These indications of specifically Syrian deviations in the belief and practice of Paul of Samosata are no more than hints (though often claimed as
. HE , .
. HE , ; the connection is indicated by J. Daniélou and H. Marrou (n. ), .
For the documentary record of the confutation by Origen of another local heresy, almost
certainly Arabian also, see J. Scherer, Entretien d’Origène avec Héraclide ().
. HE , , .
. See A. Vööbus, History of Asceticism in the Syrian Orient I (), .
. See Cypr., Ep. .
. HE , , –.
. See I.-M. Dalmais, ‘‘L’apport des églises Syriennes à l’hymnographie chrétienne,’’
Orient Syrien  (): . Note the significant generalisation (made without reference
to Paul), : ‘‘la grande Église doit s’être toujours défiée des chants aussi capiteux, trop
éloignées de cette ‘sobre ivresse’ et de cette réserve qui furent toujours siennes. . . . Au cours
du IIIe siècle une nette réaction se fait sentir en faveur de l’emploi exclusif des psaumes et
des cantiques scriptuaires.’’

Rome and the East
something more definite than that).181 But they may serve to remind us of
just how little we really understand. We can define some of the elements in
the endlessly complex culture of the Fertile Crescent in the Roman period,
and accept the impossibility of making simple deductions from culture to
political attitudes; we can confidently dismiss from the history books the
monstrous figure of ‘‘Paul of Samosata, the ducenarius of Zenobia’’; we can
see that we do not necessarily need to look to the expansion of Palmyrene
power in order to explain why the orthodox party in Antioch could appeal to
a pagan emperor. But we still are a long way from understanding the nature
of the wider Aramaic-Greek culture of Syria and Mesopotamia, and how it
affected the attitudes and beliefs of those who grew up in it.
. See J. Daniélou and H. Marrou (n. ), : ‘‘Il est typiquement oriental. On trouve
chez lui les usages de la Syrie de l’est.’’
 
Caravan Cities:
The Roman Near East and
Long-Distance Trade by Land *
My title, ‘‘Caravan Cities,’’ is intended to recall the evocative book by Rostovtzeff, published in .1 In it he gave a vivid sketch of some of the wonderful
remains to be seen then—and to be seen even better now—in the huge area
of the Near East ruled by Rome: Petra and Gerasa ( Jerash) in Jordan; Palmyra in the desert and Dura-Europos on the Euphrates, both in Syria. To say
that these were ‘‘caravan cities’’ is to say something more than merely that
trade passed through them. It is to say that they owed their character as cities
to trade: not to local exchanges of products, but to long-distance trade, to
caravans. Given the overtones of the word ‘‘caravan,’’ there is a rather strong
implication that we are talking about camel trains, travelling vast distances
over the desert.
In the case of Palmyra in particular, there has been a tendency to think
of such cities, quite explicitly, as having been like ports; hence Javier Teixidor’s book of  called the city ‘‘un port romain du désert,’’ while Ernest
Will’s of  was subtitled ‘‘La Venise des sables.’’ 2 Given the keen interest
with which we await Geoffrey Rickman’s study of Roman ports, it seemed
appropriate to use the analogy expressed in these titles and to ask how far it
* First published in M. Austin, J. Harries, and C. Smith, eds., Modus Operandi: Essays in Honour of Geoffrey Rickman (BICS supp. , ), –. An earlier version of this paper was
the subject of the Fourth Annual W. Kendrick Pritchett Lecture at Berkeley, which I had
the honour to give in April .
. M. I. Rostovtzeff, Caravan Cities, trans. D. Talbot Rice and T. Talbot Rice ().
. J. Teixidor, Un port romain du désert. Palmyre et son commerce d’Auguste à Caracalla (Semitica , ); E. Will, Les Palmyréniens. La Venise des Sables ().


Rome and the East
is justifiable to see any of the cities of the Near Eastern provinces as having
been like ports, functioning as centres for long-distance overland trade with
the Asian land-mass.
As will become clear below, there are immense problems in the evidence,
and many questions which we would like to answer will remain unanswered.
But we do have one absolutely secure starting point, namely the incontrovertible evidence provided by the Periplus Maris Erythraei (Voyage around the
Red Sea) as regards long-distance trade by sea in the zone represented by the
Red Sea, the east coast of Africa, the northern coasts of the Indian Ocean,
and the west coast of India. This highly evocative Greek text, written between ..  and , and excellently edited by Lionel Casson,3 provides just
what we do not know, or hardly know, of land trade with Asia: the various
different routes of exchange, the political relations involved, and above all
detailed accounts of the objects of trade. But it does none the less serve to validate the starting point of this essay, simply by confirming that long-distance
trade, as an organised, conscious activity, was a feature of the society and
economy of the Roman imperial period.
It would of course be very helpful if we had available any comparable work
on trade by land. In fact we do not, though the Parthian Stations of Isidorus
of Charax, apparently dating to around the turn of the eras, has sometimes
been interpreted as if it were. Indeed in his invaluable edition and translation of , W. H. Schoff sub-titled his book ‘‘an account of the overland
trade route between the Levant and India in the first century ..’’ 4 As we
will see, the work is indeed relevant for an understanding of trade routes.
But in reality the focus of interest of the author is not social or economic,
but political and military. The work begins with the crossing of the Euphrates at Zeugma, in an easterly direction, and hence (implicitly) with entry
into Parthian territory, and it ends with Alexandropolis, the metropolis of
Arachosia: ‘‘as far as this the land is under the rule of the Parthians.’’
The itinerary traversed thus comes nowhere near India, ending instead
where Parthian rule stopped in central Asia. In between, the author’s interest
is focused on local political formations, and in particular on the question of
whether a place was a Greek foundation, or counted as a Greek city. Hence
we find here one of the very few literary references to Dura-Europos: ‘‘next is
. L. Casson, The Periplus Maris Erythraei: Text with Introduction, Translation and Commentary ().
. Text of the Σταυθμοὶ Παρθυκοί in C. Müller, Geographi Graeci Minores I (), ff.;
Jacoby, FGrH, no. ; see esp. W. H. Schoff, Parthian Stations of Isidorus of Charax: An Account
of the Overland Trade between the Levant and India in the First Century .. ().
Caravan Cities

Doura of Nicanor, a foundation of Macedonians, but (which) is called by the
Greeks Europos.’’ Just before this, Isidorus has revealed what sort of travellers
his work really concerns. Speaking of the confluence of the Chabur and the
Euphrates, he says, ‘‘from this point the forces cross over to the Roman side.’’
It is probable, though nowhere made explicit, that Isidorus is thinking of a
Roman army invading, rather than of Parthian troops in transit. However,
either interpretation is possible.
The first part of the itinerary involved has been excellently discussed by
Chaumont and Gawlikowski;5 this section went across the Euphrates, continued to the Balikh and then went down this river to the Euphrates, crossing
over at the confluence with the Chabur, and continuing along the river as far
as the King’s Canal and the traverse across to Seleucia on the Tigris.
After that, the route took the traveller up through the Zagros Mountains
to Media, Margiana, and finally Alexandropolis. The terminal point is not
India, but Alexandropolis, which, as is argued by Fraser in his study of Alexander’s city foundations, can be identified with Kandahar in Afghanistan.6
It should be stressed also, however, that this city equally does not lie on any
supposed line of a ‘‘Silk Road’’ carrying trade between the Mediterranean
area and China; on any reconstruction such a route would have run well to
the north.7
The Parthian Stations, in short, is not in any simple sense evidence about
trade routes at all, still less about the legendary Silk Road. Whether there is
indeed evidence for any such route will be touched on briefly below. For
the problem of the possible role of cities lying in the Roman provinces, it is
enough for the moment to take the hint that any traveller going eastwards
might first travel south-eastwards, to the area where the Euphrates and Tigris
come closest. In which direction trade then typically ran, whether northeastwards to Media or further south-eastwards to the head of the Gulf, remains to be discussed.
The absence of any true parallel to the Periplus Maris Erythraei, giving an
account of long-distance land trade, is a serious handicap. Any such account
. M. L. Chaumont, ‘‘Etudes d’histoire parthe. V. La route royale des Parthes de Zeugma
à Séleucie du Tigre d’après l’Itinéraire d’Isidore de Charax,’’ Syria  (): –;
M. Gawlikowski, ‘‘La route de l’Euphrate d’Isidore à Julien,’’ in P. L. Gatier, B. Helly, and
J.-P. Rey-Coquais, eds., Géographie historique au Proche Orient (Syrie, Phénicie, Arabie, grecques,
romaines, byzantines) (), –.
. P. M. Fraser, Cities of Alexander the Great (), –.
. For maps of possible routes of the ‘‘Silk Road,’’ see, e.g., H. W. Haussig, Die Geschichte
Zentralasiens und der Seidenstrasse in vorislamischer Zeit (), end maps; R. Stoneman, Palmyra
and Its Empire (), . See also n.  below.

Rome and the East
would have been invaluable, whether it concerned the northern crossing of
the Euphrates at Zeugma, or that further south via Beroea and Hierapolis,
or that traversing the Syrian steppe to Palmyra and then to the Euphrates;
or possibly, much further south, that from Bostra via the Wadi Sirhan to
Dumatha and then the gulf, or that from Gaza or Petra to the Hedjaz and
south to the Yemen. As we will see, there is some explicit evidence, from
a variety of sources, for all these routes. But between them these scattered
allusions still leave two enormous problems. Firstly—unlike the Periplus—
they say extraordinarily little about what was actually carried, and what the
objects of long-distance trade were. Secondly, they leave open the question
of the importance of long-distance trade in the life of any of the cities concerned: were any of them in fact ‘‘caravan’’ cities?
If the answer is difficult everywhere else, surely it ought at least to be
easy in the case of Palmyra, lying away out in the steppe between Emesa or
Damascus and the Euphrates. Its position seems to mean that we have no
option but to conceive of it as a caravan city. As we will see, Appian, writing
in the middle of the second century, duly ascribes to it a central role in the
trade between the Roman and Parthian empires. But, quite apart from that,
we have a long series of inscriptions—local documents from Palmyra and
its region, in Greek and Palmyrene (a dialect of Aramaic)—which refer to
the caravan trade. These inscriptions stretch from ..  to the s. What
more do we want? In fact we have some more, namely the vast tax law of
.. , also inscribed in Greek and Palmyrene, and giving details of the tolls
on goods coming into the city. But both of these bodies of evidence prove,
when looked at more closely, to be somewhat deceptive.
There is a wider problem, however, indeed what seems to be a profound contradiction between what ancient sources suggest, and what modern theory dictates. By ‘‘modern theory’’ I mean above all Moses Finley’s The
Ancient Economy.8 The lessons of this book, as need hardly be said, were profound. To summarise grossly, Finley argued that neither trade as an activity
nor traders as a class ever represented a dominant factor in the life of any
ancient community. A political community might regulate trade, protect it,
or tax it: it might also seek access to raw materials, or supplies, especially
grain. But it did not have the promotion of its own trade—the search for
markets for the products of its economy—as an objective. Trade was, to use
an over-used phrase once again, ‘‘embedded’’ in a wider political structure.
One very problematic side-effect of the presentation of this sweeping hypothesis—which is in broad terms correct—was an anti-empirical tendency,
. M. I. Finley, The Ancient Economy 2 ().
Caravan Cities

that is, a disposition to take any one item of evidence, especially archaeological evidence demonstrating the movement of goods, and then reject it
as not proving enough. By rejecting each item separately, Finley did come
rather close to guaranteeing the impregnability of his main hypothesis. For
if no item of concrete evidence were allowed to mean much, no accumulation
of unexpected or unfamiliar items of empirical evidence could ever occur.
Consequently, one could never be led by the accumulation of empirical evidence into revising one’s initial hypothesis.
In the course of talking about the ‘‘ancient economy,’’ by which he meant
the economic characteristics of the Graeco-Roman world, Finley did not say
a great deal about the eastern trade of the Roman Empire. It is significant
that the Periplus Maris Erythraei is never referred to. When trade with Asia
is mentioned, it is in sweeping terms: ‘‘To be meaningful, ‘world market,’
‘a single economic unit’ must embrace something considerably more than
the exchange of some goods over long distances; otherwise China, Indonesia, the Malay Peninsula and India were also part of the same unit and
world market. One must show the existence of interlocking behaviours and
responses over wide areas’’ (). Certainly, if only some notion of a ‘‘world
market’’ or ‘‘single economic unit’’ will allow us to attribute any importance
to long-distance trade, we shall have to give up at once. For, given the restricted means of transport, and the non-existence of credit-transfer systems,
operating at a distance, without cash, it is inconceivable that long-distance
trade could have had, overall, anything like the importance which would
have produced world-wide rhythms of boom or recession of the sort familiar today. But that is hardly a reasonable requirement. What we need to ask is
two more modest questions. Firstly, did people perceive long-distance trade
with the East as an important phenomenon? Even Finley could not properly
deny that this would be a significant question: for on the very same page he
states that his justification for speaking of ‘‘the ancient economy’’ lies in its
‘‘common cultural-psychological framework’’—or, in other words (to put it
crudely), what people say in those ancient texts which happen to have survived. Secondly, long-distance trade, whose actual existence cannot possibly
be denied, might have been sufficiently important to give a different character to certain places: in short to have produced, exceptionally, some ‘‘caravan
cities.’’ Oddly enough, the Ancient Economy makes only one passing allusion to
Palmyra, and then brushes it aside as an exception: ‘‘exceptional cities, such
as . . . the caravan-city of Palmyra’’ (). But in fact, as will be seen, Palmyra
conformed rather better to his own model than he himself recognised.
If one is going to reject an approach based on the accumulation and analysis of concrete items of evidence, and look instead for a ‘‘common cultural-

Rome and the East
psychological framework,’’ one must then take seriously what the ancient
sources say. For what they (along with epigraphic evidence and iconography)
say is our means of access to ‘‘the common cultural-psychological framework.’’ But Finley wants, indeed needs, to have it both ways. Take for instance
what he says about the two passages of Pliny the Elder which reflect concern about the expense of Roman trade with the East: ‘‘The famous passages
in the elder Pliny (Natural History , ; , ) giving dubious figures of
the drain of Roman gold and silver to India and other eastern countries in
payment for luxuries are moral in their implication’’ (). But, firstly, moral
concerns are part of the ‘‘cultural-psychological framework’’; and, secondly,
moralising implications do not automatically deprive reports of all factual
content. What does Pliny the Elder, writing in the s of the first century,
actually say?
The first passage (NH , ) refers, like the Periplus, to the sea trade between Egypt and India. Each year, Pliny says, no less than  million sesterces
was drained off by India, which sent back goods sold at  times their original value. The focus of the second passage (NH , ) is the same: the luxury
trade with India, China, and the Arabian peninsula cost not less than 
million sesterces per year. There is simply no way of knowing whether the
figures are ‘‘dubious’’ or not. Both passages are indeed alarmist, or moralising:
but neither in fact mentions anything about gold or silver, each giving the
alleged values in terms of sesterces. None the less, we cannot simply brush
them aside. However curious its concerns may seem to us, Pliny’s Natural
History is the most intense exploration of man’s relation to his physical environment, in the widest sense, known to us from antiquity.9 If he expresses
concern about the cost of luxury trade with the East, then the existence of
this luxury trade was known in Rome—it was not in other words a merely
local phenomenon—and it was felt to be an issue of some importance.
It may none the less be significant that in Pliny’s eyes it was the sea trade
which was the significant factor. It is surely striking that when he first refers
to Palmyra he emphasises the richness of its soil and its water supply, and
(rather misleadingly) its position as balancing politically between the Parthian and Roman empires (NH , ); in fact—as has long been clear—it
had been firmly in the Roman sphere at least since the reign of Tiberius.
Pliny does go on immediately to give the distances from Palmyra:  Roman
miles to Seleucia ‘‘of the Parthians,’’  to the nearest point on the Syrian
. For Pliny, see the excellent study by M. Beagon, Roman Nature: The Thought of Pliny
the Elder ().
Caravan Cities

coast (a significant point, as we will see below), and  less than that to Damascus (which raises the question of whether Damascus was in any sense a
caravan city). In fact, however, even in this context, Pliny makes no reference
to trade.
The question of how ancient writers speak about long-distance trade remains, however inadequate their accounts are, of central importance. To review what we know, I would like to begin in the South, with routes through
Arabia, then go to the North, to the Fertile Crescent and Mesopotamia; and
finally come to the only area which presents a mass of local, documentary
evidence, namely Palmyra and the Euphrates.
There is a good reason for beginning in the extreme South, for it is here
that Pliny gives the most concrete and detailed available account of any longdistance trade route by land. What Pliny is describing is a route running
north along the east side of the Red Sea, starting from Thomna (or Thumna)
in the territory of the Gebbanitae (present-day Yemen) and going all the way
to Gaza on the Mediterranean coast. This is what he says:
It (i.e. Frankincense) can only be exported through the country of the
Gebbanitae, and accordingly a tax is paid on it to the king of that people
as well. Their capital is Thomna, which is ½ miles distant from
the town of Gaza in Judaea on the Mediterranean coast; the journey
is divided into  stages with halts for camels. Fixed portions of the
frankincense are also given to the priests and the king’s secretaries, but
beside these the guards and their attendants and the gate-keepers and
servants also have their pickings: indeed all along the route they keep
on paying, at one place for water, at another for fodder, or the charges
for lodging at the halts, and the various tolls; so that expenses mount up
to  denarii per camel before the Mediterranean coast is reached; and
then again payment is made to the customs officers of our Empire.10
Pliny is being quite up to date, for Gaza had probably become a part of the
province of Judaea only after the Jewish War, which had ended in .11 What
he says about the total distances is in truth unreliable, for the manuscripts
vary. But we know the distance, as the crow flies, anyway: it is approximately
, miles. If Pliny has indicated the number of stages correctly, each of the
 stages will have covered on average  miles, not allowing for deviations.
What is crucial is the indication that the goods were carried on camels,
. NH ./–, Loeb trans.
. F. Millar, The Roman Near East,  ..–..  (), –.

Rome and the East
and that successive tolls and charges had to be paid en route and finally to the
Roman publicani (tax collectors) at Gaza. Unfortunately Pliny does not give
an estimate of the original value of each load in denarii.
It is a great pity that Pliny also fails to indicate that the latter stages of this
route, before it reached Gaza, will have passed through the territory of the
kingdom of Nabataea. But we know that it must have done so, until that
area became in  the Roman province of ‘‘Arabia.’’ For the regal period, at
any rate, we can fill in the picture from the description in Strabo’s Geography,
written under Augustus and Tiberius. According to him there was a regular
trade route from Petra to Leuke Kome, at the mouth of the Gulf of Aqaba,
‘‘to which and from which camel traders travel safely and easily on the route
from and to Petra, in such numbers of men and camels as to differ in no way
from an army.’’ 12 Strabo is in fact touching here on a different aspect of the
movement of trade, namely that which came by sea up the Red Sea as far
as Leuke Kome, where goods were landed and taken up to Petra. Hence the
Periplus mentions Leuke Kome, ‘‘through which there is a way inland up to
Petra, to Malichus, king of the Nabataeans. This harbour also serves in a way
the function of a port of trade for the craft, none large, that come to it loaded
with freight from Arabia. For that reason, as a safeguard there is dispatched
for duty in it a customs officer to deal with the (duty of a) fourth on incoming merchandise, as well as a centurion with a detachment of soldiers.’’ 13
In spite of much argument, it can be taken as certain that the ‘‘centurion’’
was in fact a Nabataean officer, and the king was Malchus II, who reigned
from ..  to . Here, also, for the first time, we have a description of a
regular ‘‘caravan.’’ Ships still came up this route three centuries later, perhaps
indeed even further up, to Aqaba or Elat at the head of the gulf, the ancient
Aela. Eusebius identifies this place as ‘‘lying next to the Red Sea, sailed by
those coming from Egypt and those from India.’’ 14
From Leuke Kome, caravans will clearly have travelled up the east side
of the Gulf of Aqaba; from Aela onwards they could either go up what was
later (after the Roman annexation of the area in ) to be the Via Nova
running north-eastwards to Petra and beyond; or perhaps they may have diverged almost due north, to reach the Mediterranean at Gaza. In fact it is
perhaps more likely that even goods eventually bound for Gaza normally
went through Petra, avoiding the more mountainous parts of northern Sinai;
if so, the route will then have crossed the Wadi Arabah westwards and then
. Strabo, Geog. .. ().
. Periplus , trans. Casson.
. Eusebius, Onomastikon, ed. Klostermann, .
Caravan Cities

traversed the Negev. At any rate Strabo briefly describes the nature of the
route of  stadia between Aela and Gaza: over desert and sandy regions,
and traversed (once again) by camels.15 Elsewhere he indicates a slightly different route for goods landed at Leuke Kome: from there to Petra and then
across to Rhinocoloura (on the coast south-west from Gaza).16
Gaza seems to have played the major role as the post through which highvalue goods, such as frankincense, finally reached the Mediterranean. Roman publicani were thus stationed there in the later first century, as we know
from Pliny.17
Strabo does indeed imply, in the last passage quoted, that much of the sea
trade from the Indian Ocean had already shifted to Egypt, which is likely to
be true, and to have remained so throughout the Empire. But what needs to
be stressed is that there was a series of land routes, or combined land-and-sea
routes, from the East to the Mediterranean. If we were looking for ‘‘caravan
cities,’’ we might be in danger of finding too many. So, for instance, Strabo
also implies that in the area of Damascus there were robbers based on Trachonitis—that extraordinary broken lava field which lies like a pancake on
the plain between Damascus and the Hauran—and that, until repressed in
the later s .., they would raid merchants (emporoi) coming from Arabia
Felix.18 If that is true, it must mean that some traders, coming either by land
or sea from the Yemen, continued north from Petra, presumably along the
line of the later Via Nova, past the Decapolis region to Bostra, and on to
Damascus. If so, not two but perhaps three trade routes may have met at Damascus: namely that just implied, from the south; the road westwards over
Anti-Lebanon, into the Beqa valley and over Mount Lebanon to the coast at
Berytus; and that running north-eastwards along the line of the later Strata
Diocletiana (‘‘Diocletian road’’) to Palmyra and beyond.19 Thus Damascus also
might hypothetically count as a ‘‘caravan city.’’ 20 The evidence for this, as is
obvious, is however extremely slight, and far more significance attaches to
the extensive cultivable zone which surrounds it.21
. Geog. .. ().
. Geog. .. ().
. See C. A. M. Glucker, The City of Gaza in the Roman and Byzantine Periods (BAR
International Series , ), –.
. Geog. , ,  ().
. See text to n.  below.
. For a survey of the urban history of Damascus in antiquity, not considering possible
trade communications, see E. Will, ‘‘Damas antique,’’ Syria  (): –.
. See M. Dodinet, J. Leblanc, J.-P. Vallat, and F. Villeneuve, ‘‘Le paysage antique en
Syrie: l’exemple de Damas,’’ Syria  (): –.

Rome and the East
In recent years indeed, it has been supposed—on extremely slight evidence—that the economic centre of the Nabataean kingdom under its last
king, Rabbel II (–), shifted to the north, to Bostra.22 There is really
no good evidence for this. But one thing which has suggested it is the evidence—very slight but real—for a Nabataean and then a Roman presence
far down the Wadi Sirhan to the south-east, namely at Duma or Dumatha.
This is present-day al-Jawf, some  kilometres from Bostra, where there
are both Nabataean and Roman graffiti.23 Was there a trade route through
here? If there were, it would have involved a journey of the  kilometres
from the nearest Greek cities, Philadelphia or Bostra, and then nearly another  to the nearest point on the Euphrates, before the descent to the
Gulf. It is not impossible: the route in  stages from Thomna to Gaza was
longer. Moreover, here too, Pliny the Elder does describe such a route, even
if his ideas on distance are completely confused: he does however seem to
imply that one could travel from Petra via Dumatha to the river systems of
lower Mesopotamia.24 His conceptions are vague, to say the least. But we
know that Dumatha was within the Nabataean-Roman sphere of influence;
and there can hardly have been much point in going there except to continue
to the head of the gulf.
The trade route to the East which has always attracted the most attention is of course that through Palmyra to the Euphrates. But before we come
to that area, let us look further north. For in the nature of the case, except
in special conditions, the existence of the Fertile Crescent meant that most
traffic aiming for any point in the central Asian land-mass would leave the
Mediterranean coast in northern Syria, the area which had been suddenly
urbanised by Seleucus I around the year  ..: here there were the ports of
Seleucia and Laodicea, and the inland cities of Antiochia and Apamea, further up the Orontes; and further inland and to the north Cyrrhus (now just
on the Syrian side of the Turkish border), and to its south Beroea (Aleppo),
centred on its magnificent natural acropolis. These two places in fact defined the two main ways by which one might continue from Antiochia to
cross the Euphrates: either north-east through Cyrrhus to another Seleucia, usually known as Zeugma, ‘‘the bridge,’’ and then into Mesopotamia, to
reach Edessa; or on eastwards through Beroea, watered by the river Belus,
. See M. Sartre, Bostra. Des origines à l’Islam (), –; Millar (n. ), .
. For Dumatha, see Millar (n. ), .
. Pliny NH , . It should be clear enough that Pliny is totally confused about spatial relationships in this area. See, however, D. T. Potts, ‘‘Trans-Arabian Routes of the PreIslamic Period,’’ in L’Arabie et ses mers bordières I: Itinéraires et voyages, ed. J.-F. Salles (),
–.
Caravan Cities

and then north-eastwards to Hierapolis, before crossing the river. Here again,
we have crucial evidence from Strabo, though he in fact fails to understand
the point of his own description of where travellers went after they crossed
the Euphrates. He seems to be thinking of the more southerly crossing near
Hierapolis; but the essential point which he makes is that after crossing the
Euphrates travellers continued into the desert, and then kept the river (the
Euphrates) on their right for a three-day journey. The reason which he gives
is extremely important for understanding not only this route but also that
through Palmyra: that is, that the chiefs of the tent dwellers (Skenitai) living
further from the river were more moderate in the payments they asked for
and were more helpful in providing watering places for the camels.25 The
ambivalent relationship between traveller and local community thus surfaces
once again, as it does also on the route between the Yemen and Gaza.
Strabo does also mention the other crucial factor: namely, the two tributaries of the Euphrates which rise in the Mesopotamian shelf, the Balikh and
the Chabur. But he does not relate them to the route taken. This vital item
is supplied by Strabo’s contemporary, Isidorus of Charax. The point comes
from his brief but priceless itinerary of how one journeyed through Parthian
territory, the Parthian Stations, discussed above. He describes the more northerly route, from the Euphrates at Zeugma, and from there across to the river
Balikh; and then down the Balikh to its confluence with the Euphrates at
Nikephorion. Then the route went down the Euphrates past its confluence
with the Chabur; and then, as we saw earlier, it took the traveller back onto
the right bank, past Dura and on down the river as far as the royal canal,
where one crossed over to Seleucia on the Tigris.
The fact that both of these two northern routes—that through Zeugma
and that through Hierapolis—were well established in the late first century .. is important. For there is a danger of reading our scattered literary
references as if they added up to a consistent story—that is, a shift away both
from trade through Nabataea and through Palmyra in favour of a northern
route through the Fertile Crescent. But, firstly, the northern route itself—
or at least one variant of it—in fact turned south as soon as it reached the
Balikh; and, secondly, for the traveller there was always a choice at that point.
We can see this in the wonderful account of this area as it was in the fourth
century contained in the History of Ammianus. Two sections are of particular
importance. The first is the famous account of the fair at Batnae, or Anthemusia, which lay near the more northerly route between the Euphrates at
Zeugma and the headwaters of the Balikh:
. Geog. , ,  ().

Rome and the East
The town of Batnae, founded in Anthemusia in early times by a band
of Macedonians, is separated by a short space from the river Euphrates;
it is filled with wealthy traders when, at the yearly festival, near the beginning of the month of September, a great crowd of every condition
gathers for the fair, to traffic in the wares sent from India and China
[Seres], and in other articles that are regularly brought there in great
abundance by land and sea.26
Is the reference to ‘‘the Chinese’’ (Seres) an allusion to a ‘‘silk road’’ running
through the Asian land-mass? That is certainly possible. But there is also a
reference here to ‘‘the Indians,’’ and to goods coming by sea as well as by
land. What this means is explained when Ammianus describes Julian’s last
expedition, on which his army marched out in , from Antioch through
Beroea (Aleppo) and Hierapolis (hence by the more southerly route), and
then to the Euphrates. From there he crossed over to Batnae, thus apparently
travelling north-eastwards to join the more northerly route, and from there
continued to Carrhae, the ancient Harran: ‘‘From there two different royal
highways lead to Persia: the one on the left through Adiabene and over the
Tigris; the other, on the right, through Assyria and across the Euphrates.’’ 27
Julian then pretended to march towards the Tigris, but in fact followed the
other route, in fact exactly that traced in the Parthian Stations: that is to say,
down the Balikh to Callinicum on the Euphrates; on to the confluence of
the Chabur and the Euphrates, where Diocletian had established the fortress
of Circesium; then past the now deserted town of Dura; and then further
south to the King’s Canal, Seleucia, and Ctesiphon.
These two alternative ways of starting from Syria therefore represented
variants on an established means of reaching northern Mesopotamia. The
central feature of both was the step of avoiding the long curve of the Euphrates by crossing to the headwaters of the Balikh on the Mesopotamian
shelf, before turning south to travel down the Balikh to the Euphrates. In the
fourth century this area was Roman territory; but conversely the Euphrates
below the Chabur, including the deserted town of Dura, which had been in
Roman hands from the s to the s, was now Persian.
From the lower Euphrates, or from Ctesiphon on the Tigris, one could
travel down either river to the shores of the gulf and take ship for India. It
was on that shore, as a famous story in Cassius Dio records, that Trajan—
the only Roman emperor ever to reach the Gulf—had stood and watched a
merchantman setting off for India. If only he had been as young as Alexan. Ammianus , , , Loeb trans.
. Ammianus , , –., Loeb trans.
Caravan Cities

der! 28 The merchants were still there in (probably) the third century, when
the ‘‘Hymn of the Pearl’’ included in the Syriac Acts of Thomas describes it
as ‘‘the meeting-place of the merchants of the East.’’ 29 There is no reason to
think that this later ceased to be so.
There are two reasons why it has been supposed that the main trade routes
shifted northwards in the course of the imperial period (see text to n. 
above). One is the history of Palmyra, to which we will come. The other is
the treaty which Diocletian made, after conquering the upper Tigris region,
in  or . One provision in the treaty was that Nisibis, situated in eastern
Mesopotamia not far from the Tigris, should be the only place of synallagmata.30 The term should perhaps mean ‘‘contracts’’ or ‘‘agreements’’ rather
than specifically ‘‘exchanges of goods.’’ In any case, as we have seen, goods
sent from India and China were certainly being traded at Batnae, further west
than Nisibis, in the s. Similarly, that splendid geographical text of the midfourth century, the Expositio totius mundi et gentium (Description of the Whole
World and Peoples), describes how wealth was gained from trade with Persia
both by Nisibis but also by another city which may be either Amida, further
up the Tigris, or less probably Edessa: ‘‘for, receiving goods from the Persians,
they sell them throughout the whole territory of the Romans.’’ 31 But in any
case the precise pattern of these exchanges will have been altered profoundly
when Jovian in , after Julian’s death, had to cede Nisibis, Singara, and five
districts along the Tigris to the Persians. Moreover, we do not know what
routes trade had followed beyond Nisibis: directly to China along the ‘‘silk
route’’ through central Asia; or down the Tigris to Babylonia and the Gulf ?
In , after Julian’s death, Jovian had in fact brought his forces back up the
east side of the Tigris, and then across to Hatra in the Mesopotamian steppe,
and on to Nisibis.
The two alternatives are not of course mutually exclusive. What is certain
is firstly that this northern route, at least as far as the cities of the Mesopotamian shelf, was in use throughout the Roman imperial period. Secondly,
. Dio , , .
. A. F. J. Klijn, The Acts of Thomas. Introduction—Text—Commentary (), trans. of
,  (p. ).
. Petrus Patricius, fr. (FHG IV, ). See Millar (n. ), .
. J. Rougé, Expositio totius mundi et gentium. Introduction, texte critique, traduction, notes et
commentaire (Sources Chrétiennes , ), : Sunt ergo Nisibis et Edessa, quae in omnibus
viros habent optimos et in negotio valde acutos et bene venantes. Praecipue et divites et omnibus bonis
ornati sunt: accipientes enim a Persis ipsi in omnem terram Romanorum vendentes et ementes iterum
tradunt, extra aeramen et ferrum. Since Edessa is mentioned separately at the end of the same
paragraph, it does not seem that it can have been named in the first phrase quoted.

Rome and the East
it seems clear that camel trains were in use here too. One of the great discoveries of recent years has been the new archive of papyri of the s to
s which comes from the middle Euphrates and includes items in Greek,
Syriac, or both. One of them, of about , shows a man from Beth Phouraia,
near the confluence of the Chabur and the Euphrates, advising his son on the
hire of camels for the journey from Beroea (Aleppo) to Zeugma.32 The ‘‘caravan,’’ if that is what it was, will thus have been going from the southern route
to the Euphrates to the northern one—but where next is not made clear.
Thirdly, all the routes concerned were exposed to attacks from the nomads
whom Graeco-Roman writers called ‘‘Skenitai’’ (tent dwellers) or ‘‘Arabes’’
or (from the second century onwards) ‘‘Sarakenoi’’ or ‘‘Saracens.’’ A succession of ‘‘Saracen’’ kings, for instance, came to join Julian on his expedition
into Persia.33 But our most vivid impression comes from Jerome, in the Life
of Malchus. The story is set before , when Nisibis was lost to the Persians. Malchus was journeying eastwards from near Chalcis, on the edge of
the Syrian steppe south of Beroea, to see his parents who lived near Nisibis.
But along the road the ‘‘company,’’ deliberately travelling together for protection, were attacked by ‘‘Saracens,’’ whom Jerome, as a Christian, also calls
‘‘Ishmaelites.’’ Unfortunately Jerome does not make clear whether this episode is supposed to have taken place west or east of the Euphrates. At any
rate the narrative is too vivid not to quote:
Lying near the public highway from Beroea to Edessa, there is a desert
through which nomad Saracens are always wandering back and forth.
For this reason, travellers along the way group together and, by mutual
aid, decrease the danger of surprise attack. There were in my company men, women, old men, young people, children, numbering in all
about seventy. Suddenly, Ishmaelites, riding upon horses and camels,
descended upon us in a startling attack, with their long hair flying from
under their head-bands.34
The group whom Malchus had joined had clearly gathered for protection. He
uses for them the word comitatus, which could properly be translated ‘‘caravan.’’
. D. Feissel and J. Gascou, ‘‘Documents d’archives romains inédits du Moyen Euphrate
(IIIe siècle après J.-C.). III. Actes divers et letters (P.Euphr.  à ),’’ JS : no. ; cf.
H. M. Cotton, W. E. H. Cockle, and F. G. B. Millar, ‘‘The Papyrology of the Roman Near
East: A Survey,’’ JRS  (): –, no. .
. See Ammianus , , ; , , ; , . Cf. , .
. Jerome, V. Malchi  (PL XXIII, cols. –). Translated by M. L. Ewald in Early
Christian Biographies, ed. R. J. Deferrari (), ff.
Caravan Cities

We therefore come to the same conclusion as before. As places from which
groups of travellers set off across the margins of the steppe, exposed to attacks
from nomads, or at least to the exaction of tolls by them, Antiochia, Beroea,
and Hierapolis in Syria; Zeugma on the Euphrates; and Batnae, Edessa, and
Nisibis in Mesopotamia might all be seen, potentially, as ‘‘caravan cities.’’ The
facts of geography, and above all of the availability of water, must always
have meant that most traffic from the coastal area of the Mediterranean to
Mesopotamia, Babylonia, or central Asia travelled along the Fertile Crescent.
All this has been necessary to put what we know of Palmyra in proper
context. In this case we are talking about a very different geographical context. But it was not wholly different. It was not only, as Pliny knew (text
following n.  above), that Palmyra was a well-watered oasis. It was also that
it owed the existence of that oasis to its position under a chain of quite high
mountains which stretches out north-eastwards from the major mountain of
Anti-Lebanon above Damascus. In the mountainous, and not entirely arid,
zone north-west of Palmyra there was an extensive area of villages with small
temples and Palmyrene inscriptions, brilliantly studied by Daniel Schlumberger.35 The limit of the zone of  millimetres rainfall comes quite close
to Palmyra, whose date palms still benefit also from a natural spring.
Palmyra was certainly not, however, the centre of a rich, fertile countryside like that around Gerasa, which there is indeed no reason to think of as a
‘‘caravan city’’ at all. But, equally, it was also, like any other Graeco-Roman
city, the ‘‘central place’’ of an agricultural hinterland, as well as having, like
Damascus, a rich oasis immediately round it. It is these circumstances which
explain why, whatever is going to cast light on the long-distance ‘‘caravan’’
trade of Palmyra (which was a reality), it is not the famous customs law. For,
as John Matthews showed, this law in fact deals with products coming into
the city from its immediate hinterland.36 Indeed it is not, strictly speaking,
only a customs law, since traders, including prostitutes, are also taxed; it is
in fact a law on the indirect taxes which Palmyra, like any other city in the
Empire, was entitled to raise. But the text does refer to loads brought in
by camel or donkey; to animals coming in for slaughter; to bronze statues
being imported; to the sale of salt; and to sheep being brought in, either for
grazing or to be sheared. Almost nothing in the law suggests long-distance
trade in luxuries; indeed the only item which might do so is the reference
to camel- or donkey-loads of ointment, namely ‘‘myrrh.’’ An analysis of the
. D. Schlumberger, La Palmyrène du Nord-ouest ().
. J. F. Matthews, ‘‘The Tax Law of Palmyra: Evidence for Economic History in a City
of the Roman East,’’ JRS  (): –.

Rome and the East
tax law thus produces a picture of economic movements and of the city’s relationship to trade which could be true of any city. If Palmyra were indeed
a ‘‘caravan city’’ the tax law does not reveal it.
I am not going to argue that there was no sense in which this is true of
Palmyra. It does have to be admitted however that outsiders’ references to
Palmyra, and to its distinctive character as a city, are extraordinarily few. To
the best of my knowledge, not a single Graeco-Roman writer refers to the
remarkable art and architecture of Palmyra, even though we know that it
was visited by three emperors, Hadrian, Severus Alexander, and Aurelian.
Nor does anyone allude to the fact that in Palmyra alone, of all the cities
in the Empire, one could see a series of public inscriptions in both Greek
and a Semitic language, in parallel, which lasted for almost three centuries
(see chapter  in the present volume). Josephus however does give us a vivid
picture of Palmyra, which he supposed to have been founded by Solomon:
He [Solomon] advanced into the desert of Upper Syria and, having
taken possession of it, founded there a very great city at a distance of
two days’ journey from Upper Syria and one day’s journey from the
Euphrates, while from the great Babylon the distance was a journey of
six days. Now the reason for founding the city so far from the inhabited
parts of Syria was that further down there was no water anywhere in
the land and that only in this place were springs and wells to be found.
And so, when he had built this city and surrounded it with very strong
walls, he named it Thadamora, as it is still called by the Syrians, while
the Greeks call it Palmyra.37
Josephus’ impressions of the distances are much too short: Palmyra will have
been almost eight days’ journey from Antioch, and six days’ from the Euphrates at Dura; the journey to Babylon will have taken something like another
twelve days. But its role as a major watering place on the route between Syria
and Babylonia is clearly brought out.
Josephus, however, says nothing specifically about trade. Remarkably
enough, the only ancient writer who does refer to the trade of Palmyra is
Appian. Describing the attack on Palmyra by the forces of Marcus Antonius
in  .., Appian says, moving into the present tense, ‘‘for being merchants
[emporoi] they bring goods from [the territory of ] the Persians and dispose
of them in that of the Romans.’’ 38 Though the urbanisation of Palmyra had
already begun in the first century .., what Appian says is evidence for his
. Josephus, Ant. , –, Loeb trans.
. Appian, BC , , –.
Caravan Cities

own time, the mid-second century. It should be stressed that we have no
other external evidence for Palmyrene trade, and no internal literary evidence (there is no Palmyrene literature), nor anything more than scraps of
Palmyrene documents on perishable materials.39
In short everything that we know about Palmyrene trade is a function of
the Palmyrene version of the ‘‘epigraphic habit.’’ What is unique about this
habit is simply that it did remain bilingual, from the later first century ..
to a few years after Aurelian’s reconquest (i.e., to the late s). After that,
it is essential to stress, Palmyra was not destroyed but continued to exist as
a minor Greek provincial city, with a Roman garrison.
The inscriptions which refer to Palmyrene trade stretch from ..  to
the s.40 The first thing to stress is that, as proper off-shoots of the GraecoRoman ‘‘epigraphic habit’’ the relevant inscriptions are all, without exception, honorific inscriptions for individuals. The earliest of them comes from
.. , put up by Greek and Palmyrene merchants ‘‘in Seleucia’’ (BSLWKY’
in the Palmyrene text, in fact the only reference to Seleucia in the epigraphy
of Palmyra).41 The next is of .. , put up to honour Malichos son of Nesa
by ‘‘the merchants who are in Babylon, because he benefitted them in all
ways.’’ 42 The series continues to the s, with a marked concentration in the
earlier second century. The latest firmly dated one is of / and is put up
by the boulē and dēmos (council and people) of what was now a Roman colony
to honour Iulius Aurelius Salamallathus, an archemporos (leading merchant),
for having ‘‘brought back’’ a caravan (synodia) ‘‘without charging, at his own
expense.’’ 43 The second thing to note is that all the places mentioned in the
inscriptions lie to the east of Palmyra: the city of Seleucia on the Tigris is
mentioned, as we have just seen, only once, on a fragmentary inscription of
.. . Otherwise we have references to Vologaesias, somewhere on the Euphrates, to Mesene at the head of the Persian Gulf, and, very significantly, to
‘‘Scythia.’’ We know what they will have meant by ‘‘Scythia,’’ because the Periplus () also refers to this area—it was north-west India. So, for instance, one
of the many inscriptions of the mid-second century in honour of M. Ulpius
Iaraios, or Yarhai, was put up ‘‘by the merchants who have returned from
. For the known fragments, from Palmyra itself and Dura, see Cotton, Cockle, and
Millar (n. ), nos. , ; the latter is listed also as PAT, no. .
. All the known inscriptions relating to Palmyrene trade are conveniently listed by
M. Gawlikowski, ‘‘Palmyra as a Trading Centre,’’ Iraq  (): –, on –. Those referred to in this paper will have their numbers in PAT added.
. CIS II., no.  Inv. IX, no. A PAT, no. .
. CIS II., no.  Inv. IX, no.  PAT, no. .
. CIS II., no.  Inv. IV, no.  PAT, no. .

Rome and the East
Scythia in the ship of Onainos son of Addoudanos,’’ because of the assistance
which he had given them.44
Palmyrene traders, it is quite clear, regularly travelled across the steppe
to the Euphrates, either directly to Dura or on the route south-westwards
to Hit (Aeipolis). The long track across the steppe is still visible from the
air. They might go as far down the river as Vologaesias, whose exact location on the Euphrates is not known, or on to Spasinou Charax (Mesene),
which is mentioned several times in the inscriptions from the second century. But they might also take ship to north-west India. Strictly speaking
there is no Palmyrene evidence to show that they ever took the route northeastwards into central Asia, starting from Seleucia and crossing the Zagros
Mountains, which Isidorus’ Parthian Stations indicates (see text to n.  and
to nn. – above). As Gawlikowski has made clear, the Palmyrene trade
route reflected in the inscriptions runs, without exception, south-eastwards
towards the head of the gulf: ‘‘There is strictly nothing to suggest that the
Palmyrenes were interested in the land route through Iran and Central Asia,
the celebrated Silk Road.’’ 45
There did seem at one moment to be real archaeological evidence for a
Palmyrene presence on the Silk Road, in the form of two Palmyrene stēlai
(inscribed pillars) found at Merv in present-day Turkmenistan; but unfortunately it now seems clear that they were brought there in the nineteenth
century.46
What we can know from the honorific inscriptions is confined—obviously enough—to what they will tell us. Firstly, there were indeed ‘‘caravans’’—synodiai. Their members are called in Palmyrene BNY ŠYRT’ (sons of
the caravan). Secondly, these caravans required protection against the peoples
living in the steppe. The protection involved might simply be financial, in
other words the payment of something between tolls and protection money
(compare Pliny’s account of the route through the Hedjaz, and Strabo’s of the
Euphrates route: text to nn. – and – above). We have seen how Salamallathos in  brought back a synodia ‘‘at no cost,’’ ‘‘at his own expense.’’ But
on other occasions real fighting was involved. In  Ogelos son of Makkaios
is honoured ‘‘for his having given satisfaction through continued commands
[stratēgiai ] against the nomades, and having provided safety for the merchants
[emporoi] and the caravans [synodiai] in all his commands of caravans [syno. Inv. X, no.  PAT, no. .
. Gawlikowski (n. ), .
. For the evidence and references, see K. Parlasca, ‘‘Auswärtige Beziehungen Palmyras
im Lichte archäologischer Funde,’’ Dam. Mitt.  (): –, on .
Caravan Cities

diarchai].’’ 47 We know that the Palmyrenes, uniquely for a provincial city,
maintained military outposts far away on the Euphrates, outside the Roman
provincial area.48
Our knowledge of the difficulties and dangers which faced caravans on
their way back from Vologaesias to Palmyra has been enhanced by a recently
published bilingual inscription of ..  from the temple of Athena, honouring Soados son of Boliades (to use his Greek name). He had saved a caravan under attack from forces under a leader called in Palmyrene ‘Abdallat
Ahiitya, and the allusion here seems to be quite clearly to military protection.
Since the preliminary publication appears in an article by H. J. W. Drijvers
on the dependence of the public vocabulary of Palmyrene on Greek,49 and
might thus escape notice altogether, it is worth reproducing Drijvers’ translation here, with minor presentational changes:
[The statues?], this one in [the temple of Athen]a, one in the sacred
garden, and one in the temple of Atargatis, [which] have been erected
next to the first four statues that were erected by the first caravan [synodia] for Soados son of Boliades son of Thaimisamsos, who is pious and
patriotic, through his benevolence and magnanimity towards the citizens in every way, adorned with distinctions and very great honours,
the caravan [synodia] of all Palmyrenes which came back from Vologaesias erected, because he advanced in a distinguished manner, taking
with him a large force, and he protected (them) against [Ab]dallathos
from Eeithe and these [‘‘the robbers’’ in the Palmyrene text] that were
brought together by him from P[.......] who for a long time were lying
in wait in order to harm the [caravan...] he preserved them. Therefore they erected for him [the statues?] to honour him, when Malê
son of Sumonos [.....] and [E]nnibel son of Sumonos son of Bazekes
were caravan-leaders [συνοδιαρχούντων] in the year  month Daisius
( June).
The Abdallathos against whom Soados had protected the caravan will presumably have been very like the phylarchoi (tribal chiefs) of the Skenitai of
whom Strabo speaks as operating on the other side of the Euphrates (text
. Inv. X, no.  PAT, no. .
. See M. Gawlikowski, ‘‘Palmyra et l’Euphrate,’’ Syria  (): –; Millar (n. ),
, –.
. H. J. W. Drijvers, ‘‘Greek and Aramaic in Palmyrene Inscriptions,’’ in Studia Aramaica:
New Sources and New Approaches, ed. M. J. Geller and J. C. Greenfield (), –, on –,
with pl. . Not included in PAT.

Rome and the East
to n.  above). This seems to be the only occasion on which one of them is
mentioned by name in the epigraphy of Palmyra.
We should not minimise what we can learn from these inscriptions, limited as they are in their formal character. Synodiai, ‘‘caravans,’’ were indeed a
regular feature of trade. There was a public position called synodiarchos (commander of a caravan) and another called archemporos (leading merchant). These
caravans travelled across the steppe to and from Vologaesias and Spasinou
Charax. Some Palmyrene merchants ventured further, across the Indian
ocean. Protection against the dangers of the land route was necessary and
was supplied either by money or by actual fighting against the nomades.
But we learn nothing of the internal organisation of a synodia; nothing
whatsoever of the objects of trade; nothing directly about the forms of transport, presumably camels or donkeys; and above all we hear nothing of how,
or in what direction, this trade was then carried westwards—that is, of how,
in Appian’s words, the Palmyrenes disposed of the goods in the empire of
the Romans (see text to n.  above).
What this means is that what the inscriptions will tell us is determined
by their political and, to an extent, military preoccupations: the benefits
which the euergetai (benefactors) who were honoured conferred related to
the no-man’s-land of the steppe between Palmyra and the Euphrates, and
between the Roman and the Parthian, and then the Persian, empires. The
routes to the west, towards the coast, will have been relatively secure. One
certainly led south-eastwards to Damascus; by the end of the third century
this route had an imperial name, the ‘‘Strata Diocletiana,’’ and was lined with
Roman forts.50 Anything which reached the Mediterranean from there will
have been carried through the pass over Anti-Lebanon into the Beqa valley
and then over Mount Lebanon to Berytus. Much more direct was the route
westwards to Emesa (Homs), or Laodicea ad Libanum, and then through the
Homs gap to the sea. When traffic reached the sea, the obvious ports were
the old Phoenician cities of Arados to the north and Tripolis to the south.
But to say that is immediately to reveal how little impression we have of the
trading role of these very modest places under the Empire. There is no doubt
of course that cargoes could be picked up by trading ships moving along this
coast either southwards, to Egypt, or more importantly northwards and then
along southern Asia Minor to the central Mediterranean. Saint Paul, on his
last missionary journey, took a ship going from Patara southwards to Tyre,
Ptolemais, and Caesarea (Acts :–). Then, when sent to appear before
. See Millar (n. ), –; see T. Bauzou, ‘‘Epigraphie et topographie: le cas de la
Palmyrène du sud-ouest,’’ Syria  (): –.
Caravan Cities

Caesar, he was put on another going northwards, from Caesarea to Sidon,
Cyprus, and Myra, where he was transferred to an Alexandrian corn ship
bound for Rome (:–). The Pseudo-Clementines also contain a quite detailed account of a voyage up the Phoenician coast, stopping at a number
of places.51 We do not need to multiply testimonies, for it is beyond doubt
that, once at the coast, goods could be transferred to the Mediterranean sea
trade. High-value, low-volume cargoes could easily be put on such ships as
Acts describes. But have we any concrete evidence that such cargoes, transported from the interior, were in fact traded on the Phoenician coast? The
answer is somewhat embarrassing, namely a single passage of Galen. It comes
in his description of how he collected specimens of chemicals and drugs on
a journey to Lemnos, Cyprus, and Palaestine:
Moreover I had the good fortune to lay hold of the Indian lycium which
had been recently imported to Phoenicia together with the Indian aloe.
This happened when I was on my way back from Palaestina, and I
was persuaded that the lycium was Indian both by the very fact that
it was brought by camels, together with the whole cargo and because
the spurious one could not be known to those who brought the lycium as the material from which it is prepared is not produced in their
localities.52
Galen makes his reasoning quite clear: the Indian product, lycium, could be
regarded as being both authentically Indian and as having been imported
overland, since the whole consignment had been brought by camels ( phortion here surely refers to the presumed transport by land). The carriers must
therefore, it seems, have come via Palmyra.
Galen’s conclusion is of some significance. For lycium is among the products which the Periplus describes as being imported by sea from north-west
India to Egypt.53 Pliny the Elder also adds that it was sent by ‘‘the Indians’’
in the skins of camels or rhinoceruses.54
Galen does not tell us where these caravans reached the Phoenician coast,
or indeed precisely where he acquired the goods. But his presumption is
enough to produce further possible candidates for the title of ‘‘caravan city’’:
Emesa, whose rise and fall, as Seyrig suggested, exactly parallels that of Pal. Pseudo-Clementine Recognitiones, Homelies –.
. Galen, de simplicium medicamentorum temperamentis, ed. Kühn, vol. XII, ; text with
translation in Stern, Greek and Latin Authors II, no. , whose translation is used here.
. See Casson (n. ), esp. –, –.
. Pliny, NH , .

Rome and the East
myra to its east;55 perhaps even Laodicea ad Libanum—if geographic determinism really determined anything, and if trade had been a major factor in
the growth of cities, this place ought to have been one of the major cities of
the ancient world. For it lies in the Homs gap between Palmyra and the sea,
and just where that route crosses the Orontes as it flows north from the rich
Beqa valley, towards Apamea and Antiochia. The vast tell of ancient Qadesh
shows that in the past it had indeed been a major city. But in the Roman
Empire it was not.
If we survey what we have found, bearing in mind how slight our literary evidence is, so that negative arguments—suggesting the non-existence
of trade routes—could never rest securely on it, we find that there certainly
were well-known long-distance trade routes by land. Firstly, in the south
of the Roman Near East, it is quite clear that trade did cover the vast distance by land from the Yemen through the Hedjaz to the Mediterranean at
Gaza or Rhinocoloura. It ran in parallel with a sea trade which might bring
goods from the Indian Ocean to either the Egyptian or the Arabian shore
of the Red Sea. Goods brought originally by either means might well pass
through Petra. But much more evidence would be needed before we could
properly characterise Petra as a ‘‘caravan city,’’ rather than as a royal capital,
and then modest provincial city, situated in a zone where agriculture was
possible, and a number of minor settlements also existed. The emerging evidence of the newly discovered sixth-century papyri from Petra makes this
agricultural hinterland visible for the first time.56
Rather further north, it is probable that a trade route ran south-eastwards
from the Hauran and Bostra through the Wadi Sirhan to Dumatha and then
to the gulf. But we have so far no idea of the volume or nature of any such
trade, or whether it will have had any effect on the character of cities in
the provincial area. Just conceivably, trade from this direction, or that continuing northwards from Petra, may have passed through Gerasa, whose fine
central plaza, with its unique oval shape, suggested to Rostovtzeff a possible
role as a meeting place for caravans. But given the rich territory of Gerasa,
no such explanation for the architectural grandeur of the city is required. As
with Damascus itself, where various trade routes may have met, we can only
say that the evidence, while it does not exclude, also does not require any
characterisation of it as a trading, or ‘‘caravan,’’ city.
It should be stressed emphatically that if long-distance trade by land had
been important anywhere, it should have been not in this southern zone
. See H. Seyrig, ‘‘Caractères de l’histoire d’Émèse,’’ Syria  (): –.
. L. Koenen, ‘‘The Carbonized Archive from Petra,’’ JRA  (): –.
Caravan Cities

but in the north, where traffic along the Fertile Crescent will have reached
the Mediterranean, crossing the Euphrates at Zeugma or near Hierapolis and
proceeding towards Beroea (Aleppo) and Antiochia, and then the fine ports
founded in the early Seleucid period, Seleucia and Laodicea. If we look in
the other direction, eastwards to central Asia, all the concrete indications are
that traffic followed the Euphrates and Tigris south-eastwards at least as far
as Seleucia on the Tigris, and that it was (at least) characteristic for it to continue further to Mesene. There this land trade route became a sea route, in
which ships sailed first to ‘‘Scythia,’’ the north-west corner of the Indian subcontinent, and might go further south along the coast, and on occasion to
Taprobane (Sri Lanka) or into the Bay of Bengal.
This is not the place to explore the sea trade in the Indian Ocean, illuminated now by a Greek contract on papyrus drawn up at Muziris.57 What
is important is to stress that in ‘‘Scythia,’’ as the Periplus (; ) shows, silk
and Chinese pelts were obtainable. According to the Periplus () goods from
China could be brought overland through Bactria and then to northern
India, reaching the sea either by being carried down the Ganges to the Bay
of Bengal or at Barygaza, one of the ports of ‘‘Scythia.’’
We can take it as certain, therefore, that caravans did leave the central
and northern areas of the Roman Near East, and did travel down the great
river systems to the gulf, where the sea trade to India put them in contact
with goods brought overland from China. These considerations must raise
the question of whether there was indeed a trade route following the alternative itinerary which Isidorus maps out in his Parthian Stations, from Seleucia
north-eastwards across the Zagros, to Media and Bactria—and then hypothetically continuing, as Isidorus’ route does not, eastwards to China. Finds
of Chinese products in the Mediterranean area, for instance fragments of
Chinese silk at Palmyra,58 would obviously, of themselves, not prove the existence of a central Asian route, as opposed to any other. However, it would
be absurd to deny that there were indeed a number of land routes through
central Asia, or that trade travelled along them, in both directions. The first
book of Ptolemy’s Geography shows both that such routes were known and
that they were used for trade: hence, for instance, Ptolemy’s indication of a
journey of , stades from the crossing of the Euphrates near Hierapolis
to the ‘‘Stone Tower,’’ and , from there to China.59
. See, e.g., the articles collected in Topoi Orient-Occident . (), under the title
‘‘Inde, Arabie et Méditerranée orientale.’’
. See, e.g., Parlasca (n. ).
. Ptolemy, Geog. , , .

Rome and the East
The evidence on these central Asia trade routes cannot be considered
here.60 All that needs to be asserted is that there were indeed ‘‘caravans’’ using
long-distance routes, both within the area of the Roman provinces (which in
the s came to include northern Mesopotamia) and outside it, and that they
characteristically tended to go as far as the ‘‘Kings’ Canal’’ which joined the
Euphrates and Tigris. What the relative importance was, after that point, of
the overland route north-eastwards and the land-sea route south-eastwards
cannot be determined.
Were there then ‘‘caravan cities’’? All that can be said is that there were a
large number of places through which long-distance trade will have passed.
In the case of none of them, with one exception, does the evidence require a
characterisation in terms of their trading role, as opposed to normal political
or social functions, or an economic role in relation to the surrounding territory. But to say that is categorically not to prove a negative; it is to leave
the question open.
The one exception is, of course, Palmyra. Even here, modern research has
revealed the local geographical factors which made possible the existence of
a city, as it has also the existence of an extensive hinterland where agriculture was practised. As it happens, precisely this is demonstrated by the actual
character of the famous tax law. It is also of fundamental importance that all
our extensive local evidence on the long-distance land (and sea) trade at Palmyra is due to a local variant of one of the most distinctive features of the
political structure of the actual city: the honorific inscription as a specimen
of the ‘‘epigraphic habit.’’ In that sense, even Palmyra conforms to Finley’s
general view that trade has to be seen as ‘‘embedded’’ in a political structure. For what is categorically ‘‘embedded’’ in this sense is the whole of the
documentation on which our knowledge of Palmyrene trade depends.
That said, however, the real world which the honorific inscriptions do
(partially) reveal is indeed one of caravans organised to travel across the
. A full study of this question would require a survey of archaeological finds both
along the possible routes and at either end, as well as an analysis of literary and documentary evidence relating to journeys, and to social and economic structures. For some relevant approaches, see, e.g., M. Raschke, ‘‘New Studies in Roman Commerce with the East,’’
ANRW II.. (), –; W. H. Haussig, Die Geschichte Zentralasiens und der Seidenstrasse in vorislamischer Zeit (); R. Whitfield and A. Farrer, eds., Caves of the Thousand
Buddhas: Chinese Art from the Silk Route (); Stoneman (n. ), chap. : ‘‘Of Spices, Silk
and Camels.’’ See also the the papers from an international conference on ‘‘Palmyra and the
Silk Road,’’ held in Palmyra in  and published as a special issue of Annales Archéologiques
Arabes Syriennes  (). I was very grateful to Jessica Rawson for a valiant attempt to
reduce my ignorance in this area.
Caravan Cities

steppe for purposes of trade, with merchants and ‘‘chief merchants’’ and ‘‘caravan leaders’’ taking steps to protect themselves against the nomads, either by
negotiation or by suitable payments or by actual physical force. Even Palmyra did not live only by or for long-distance trade. But it is precisely the
distinctive nature of the services for which its leading citizens might be honoured which makes it impossible to deny that Palmyra at least was indeed a
‘‘caravan city.’’
           
Looking East from the Classical World:
Colonialism, Culture, and Trade
from Alexander the Great to Shapur I *
The emperor Trajan’s arrival at the head of the Persian Gulf in ..  was
the only visit there by a Roman army in the course of the wars against the
Parthian and Sassanid empires which continued sporadically for centuries.
The occasion gave rise to a well-known and illuminating anecdote, told by
Cassius Dio in his Roman History. As it happens, both the author and his History themselves illustrate the relationship between the classical world and
Asia; for Dio was a Greek who came from Bithynia, the north-west corner of
modern Turkey, but who was also a Roman senator, and was twice consul in
the early third century. Looking back to Trajan’s campaign a hundred years
earlier, he writes in book  of his eighty-book History in Greek: ‘‘Then he
came to the Ocean itself, and when he had learned its nature and had seen a
ship sailing to India, he said, ‘I would certainly have made an attempt against
the Indians, if I were still young.’ For he began to think about the Indians,
and was curious about their affairs, and said how lucky Alexander had been.’’ 1
* First published in International History Review  (): –.
This paper was a revised and expanded version of a lecture given at the Anglo-American
Conference of Historians in London in July , and also at Macquarie University, Sydney,
in November . As will be obvious, it can claim to do no more than touch, in a selective
and personal way, on a few of the profound issues raised by this topic. References to both
ancient evidence and modern literature are also very selective, and are intended to bring
out particularly interesting recent discoveries, and to suggest further reading. For lesserknown ancient works I have noted English translations where available. I was extremely
grateful to Amélie Kuhrt for detailed comments on an earlier draft of this paper, and to
Jenny Graham for drawing the map.
. Cassius Dio, Roman History ,  (Loeb trans., vol. VIII, –). For Trajan’s campaign, see J. Bennett, Trajan, Optimus Princeps: A Life and Times (), chap. XIII.

Looking East from the Classical World

The anecdote exemplifies one aspect of the relationship between the classical world and the peoples of Asia: imperialist dreams and ambitions, and
actual conquest, whether long-lasting or ephemeral. Along with this went
an ideology of colonialism underpinning a vision of the export of Greek, or
Graeco-Roman, city life and civilisation to distant lands. These goals cause
conceptual, and even moral, problems for us, obliging us to ask whether we
can detach ourselves from a Western imperialist perspective. The difficulty
arises not only from our lack of linguistic and scholarly training in the numerous and varied cultures, languages, and scripts in use in ancient Iran, central Asia, and India, but also from the lack in some areas of ‘‘native’’ written
evidence, or the existence only of texts whose date and character is profoundly uncertain. In parts of Asia, the only ‘‘Asian’’ culture or belief system
encountered turns out to have been one constructed by Graeco-Roman observers. Detaching oneself from a ‘‘classical,’’ or Western, perspective, therefore, is more difficult than it might appear at first glance.
This essay will look eastwards from the Mediterranean. It will set out
the broad lines of Graeco-Roman military and political involvement in Asia,
from the Mediterranean through Iran to central Asia and northern India;
point to the surviving traces of Greek culture and language from these regions; and discuss for which possible trade routes to the East we have reliable
contemporary evidence for the Roman imperial period. To say that we do
not have reliable evidence for a supposed trade route is not to claim that no
such route existed. None the less, no adequate contemporary evidence survives from the classical period to demonstrate the existence of a ‘‘Silk Road’’
crossing central Asia to China.
Beyond the concrete, prosaic issues of wars and states, documents and
traders, lie more complex questions of cultural and religious history, and of
profound mutual influence. The essay does not attempt to explain, for example, the influence of Babylonian mythology and astronomy, or of ‘‘Zoroaster’’ and ‘‘Iranian dualism.’’ Nor does it explain the origins and spread of
Manichaeism, or of the preaching of Christianity in Asia. But it does examine the two remarkable Hellenistic inscriptions which represent the only
known expression of Buddhism in classical Greek.
The essay focuses on the six centuries following Alexander’s conquests.
The culture of archaic Greece, from the eighth to the sixth centuries, and
contemporary with the Neo-Assyrian, Babylonian, and Achaemenid Persian empires, emerged in the shadow of the major Near Eastern cultures,
of Anatolia, Phoenicia, Egypt, and Babylonia. Several key developments,
for instance, the derivation of the Greek alphabet from the Phoenician and
the debt of Greek conceptions of the gods to Egypt, were recognized by

Rome and the East
Herodotus as early as the fifth century ..2 The importance of Babylonian
mythology for early Greek literature, on the other hand, was unknown to
Herodotus, and has only become clear in the course of the nineteenth and
twentieth centuries, with the publication of Akkadian texts.3 Whether we
think of art, literature, astronomy, or conceptions of the world, the debt of
early Greece to ‘‘Asia’’ is accepted by everyone and is the subject of important
scholarly works.4
The ‘‘classical’’ period of Greek history, following the successful Greek resistance to the invasions by Darius and Xerxes in the Persian wars, was shaped
by political and military relations with the Achaemenid Empire. For much
of the sixth, fifth, and fourth centuries, the Persians ruled the western shores
of the Aegean, where a large number of Greek cities were located, and Persian control extended round the eastern shores of the Mediterranean as far
as Egypt and much of present-day Libya, where there had also been extensive Greek settlement. The relationship between the Greek world and Persia,
and the Greeks’ resistance to Persian conquest, was of course the subject of
Herodotus’ Histories, of Xenophon’s Anabasis and to a certain extent of his
Hellenica, and of Isocrates’ remarkable analysis of the political situation of the
Greek world as it was in  .., the Panegyricus.5
The cultural and artistic influences of the period are difficult to evaluate,
not least because the empire ruled by the Persians embraced so many contrasting cultures, from Anatolia (with its Greek cities) to Phoenicia, Egypt,
Babylonia, and their own homeland of Iran (the great mystery) itself. Although important exchanges in art and architecture occurred,6 most significant for the future was the emergence in the fifth and fourth centuries of
. Herodotus , – (‘‘Phoenician letters’’); , ;  (borrowing of divine names
from Egypt).
. Note especially the useful collection and translation, Myths from Mesopotamia: Creation,
the Flood, Gilgamesh and Others by S. Dalley (), and the much fuller presentation of the
material by B. R. Foster, Before the Muses: An Anthology of Akkadian Literature I–II ().
. Note, e.g., S. P. Morris, Daidalos and the Origins of Greek Art (); W. Burkert, The
Orientalising Revolution: Near Eastern Influence on Greek Culture in the Early Archaic Age ();
or the massive work by M. West, The East Face of Helicon: West Asiatic Elements in Early Poetry
and Myth ().
. Translated in the Loeb edition of Isocrates, vol. I, ff. This much-underestimated
text would be well worth studying as perhaps the earliest example of a systematic essay in
strategic-political analysis.
. See M. C. Miller, Athens and Persia in the Fifth Century BC: A Study in Cultural Receptivity ().
Looking East from the Classical World

a Greek ethnic, or consciously national, identity, which contrasted Greeks
with barbaroi.7 With that identity came indications of a claim to military and
political superiority. As Greeks, for instance, practised athletics naked, in the
sunlight, Xenophon records that at Ephesus in  the king of Sparta ordered
that Persian prisoners being put up for sale should be stripped, to show their
Greek captors how soft and white their bodies were.8 Such attitudes led to
a tendency to see non-Greeks in Asia as ‘‘lacking in spirit,’’ and as not in
possession of, or being incapable of exercising, political rights.9 From this
followed the perception that the Persian Empire might not be as strong as its
vast extent suggested, an argument made by Isocrates in the Panegyricus on
the evidence of the ‘‘Anabasis,’’ the ‘‘march up-country,’’ in  in support of
Cyrus’ claim to the Persian throne. Cyrus’ forces had included Greek mercenaries, among them Xenophon himself, on whom Isocrates based his claim
that the empire was more vulnerable than it seemed.10
When a new power, the Macedonian kingdom, arose in the mid-fourth
century on the northern borders of the Greek world, it was Isocrates again
who in  petitioned Philip II to take up an explicitly imperialist and colonialist role against Persia on behalf of the Greek world, partly to provide
employment for the large numbers of unsettled mercenaries who, Isocrates
believed, were destabilizing it:
What opinion must we expect the world will have of you . . . if you
undertake to conquer the whole empire of the King, or, at any rate,
to wrest from it a vast extent of territory and sever from it—to use a
current phrase—‘‘Asia from Cilicia to Sinope’’ [western Asia Minor];
and if, furthermore, you undertake to establish cities in this region,
and to settle in permanent abodes those who now, for lack of the daily
necessities of life, are wandering from place to place and committing
outrages upon whosoever they encounter? 11
This is not the place for any detailed account of how Philip’s son Alexander (r. – ..) not only fulfilled this programme but also exceeded it.12
. See especially E. Hall, Inventing the Barbarian: Greek Self-Definition through Tragedy
().
. Xenophon, Hellenica , ,  (Loeb ed., vol. I, ).
. E.g., Aristotle, Politics ,  (b).
. Isocrates, Pan. – (trans. Loeb, vol. I, –).
. Isocrates, Address to Philip,  (trans. Loeb, vol. I, ).
. For modern accounts, see A. B. Bosworth, Conquest and Empire: The Reign of Alexander
the Great (); P. Briant, Alexandre le Grand 4 ().
Looking East from the Classical World

Alexander’s conquests may be divided into two distinct phases with markedly different long-term consequences. The first phase, which itself exceeded
anything Isocrates had envisaged, may be defined as the termination of the
Persian dynasty by the conquest of the parts of the Near East with which the
Greeks were familiar—Anatolia (though not all of it at first), Syria, Palestine (hence, with momentous consequences, the small Jewish community),
Egypt, and Babylonia, which, as a prosperous settled area with an ancient culture, was one of the heartlands of the Persian Empire 13—and by the capture
of the Achaemenid ‘‘capitals,’’ Susa and Ecbatana.
There followed in phase two Alexander’s still almost entirely unintelligible expedition across Iran and central Asia to present-day Afghanistan and
Tajikistan, before turning south to the Indus Valley, and home again by way
of a gruelling march along the northern shores of the Indian Ocean and the
eastern shores of the Persian Gulf (see map).14
Both phases of conquest had important long-term effects. The first created
a ‘‘Greek’’ world in Anatolia, the Levant, and Egypt, and to a lesser degree in
Mesopotamia, which lasted for more than a thousand years. The second gave
rise to the imperialist and colonialist dreams which were to haunt the Roman
Empire from the late Republic into the Christian period, and to complex cultural interchanges, still very little understood, of which the existence of the
two Greek Buddhist texts discussed below is a pointed reminder. Alexander’s
march also marked out the part of Asia with which the classical world would
become reasonably familiar, and with which it enjoyed regular contact during the Roman imperial period. Of the area beyond Alexander’s march, little
or nothing was known.
Viewed from the Mediterranean, the part of the Near East stretching from
present-day Turkey to Egypt became, as a result of Alexander’s conquests,
effectively Greek. Greek culture was dominant, Greek cities and Greek architecture were visible everywhere, and the Greek language used both for
monumental inscriptions and perishable documents. The depth of the penetration by Hellenism, the nature of Greek settlement and colonialism, and
the character of the principal successor state, the Seleucid kingdom, which
inherited many features of the Persian Empire, have all, however, been the
. See, e.g., M. Stolper, ‘‘Mesopotamia, – ..,’’ in J. Boardman et al., eds., Cambridge Ancient History IV: Persia, Greece and the Western Mediterranean c.  to  .. (),
ff., and P. Briant, De Cyrus à Alexandre: Histoire de l’empire perse (), esp. ff.
. For this extraordinary episode, covering the years  to  .., see A. Bosworth,
Alexander and the East: The Tragedy of Triumph (), and, for the central Asian context,
F. Holt, Alexander the Great and Bactria: The Formation of a Greek Frontier in Central Asia ().

Rome and the East
subject of vigorous debate. Owing to the Seleucid monarchy’s use of official languages other than Greek (for instance, Aramaic and Akkadian), it has
recently been portrayed as a Near Eastern rather than a Greek state in a cultural as well as geographical sense.15 None the less, the dominance of Greek
forms of political organisation and of the Greek language survived the defeat
of the Hellenistic kingdoms by the Roman Republic in the second and first
centuries .., and lasted throughout the Roman imperial period, long after
, when the emperor Constantine converted to Christianity. The frontier
of the Empire lay along the middle and upper Euphrates, until it advanced
to take in all, and then (after a defeat by the Sassanid Persians in ) most,
of northern Mesopotamia.
Of course, there is evidence of the continued use of local languages in
Anatolia, and of various dialects of Aramaic in Syria, and of Hebrew and
Jewish Aramaic in Judaea. The Jewish community was the sole example, however, of a national tradition, embodied in Semitic-language texts, which
persisted throughout the Graeco-Roman period. From the third century
.. onwards, the books of the Hebrew Bible were progressively translated
into Greek, making possible the extension of Christianity to gentiles and
the writing of Christian Scriptures in Greek. The spread of Christianity in
turn promoted the emergence of Christian ‘‘languages of culture’’ other than
Greek, for instance, Coptic, Syriac, and Armenian, and of non-Greek forms
of Christianity that survive to the present day.16
During the classical period itself, with the exception of a few periods of
Jewish independence, some long, others short—the Hasmonean state, which
freed itself from Seleucid rule step by step from the s to  .. and
remained independent until  ..; the revolt of ..  to ; and the
Bar Kochba revolt of –—few non-Greek political structures were to
be found west of the Euphrates during the millennium from Alexander to
Muhammed (the Nabataean kingdom of the first century .. would be a
partial exception). In the eyes of Graeco-Roman observers, the only local
culture and religious tradition which represented anything distinctive, trou. For this thesis, note esp. S. Sherwin-White, ‘‘Seleucid Babylonia: A Case Study for
the Installation and Development of Greek Rule,’’ in A. Kuhrt and S. Sherwin-White, eds.,
Hellenism in the East: The Interaction of Greek and Non-Greek Civilizations from Syria to Central
Asia (), ; and, more fully, S. Sherwin-White and A. Kuhrt, From Samarkhand to Sardis:
A New Approach to the Seleucid Empire (), with a set of discussions in Topoi  ().
. For a useful survey, M. Albert, ed., Christianismes orientaux: introduction à l’étude des
langues et des littératures ().
Looking East from the Classical World

bling, or worthy of comment was Judaism.17 These two statements are true
even of Egypt, despite the persistence of Egyptian temples and priests, of
monumental (hieroglyphic) and cursive writing in Egyptian, and of Egyptian art and architecture (though Egypt represents too complex and individual a case in cultural history to be explained here). With few exceptions,
the Graeco-Roman observer saw the Near East as Greek, marked out by a
network of Greek cities and, from the fourth century onwards, by Christian
churches and bishoprics.
Anatolia remained Greek and Christian until the Turkish advances of,
broadly, the eleventh century. In spite of the settlement there of a Celtic
population, the ‘‘Galatians,’’ in the third century .., the use of Phrygian
for public inscriptions in the Roman period, and the persistence of Phrygian
and Lycian architectural styles, the cultural and religious history of Anatolia, even of inland areas—let alone the Greek cities along the coast—from
the Hellenistic period to the late Empire is the history of Greek culture, in
both city and country.18 The same is true, despite the visibility of Semitic
languages, to the south of Anatolia.19
To think, when looking at the Near East, of the classical world and Asia
as two different worlds, or of the latter as a foreign, ‘‘Asian’’ world as seen
from the Mediterranean, is entirely to misconceive the situation. West of the
Euphrates, and in some areas east of it, the Near East was a part of the classical world. Our contemporaries who make historic claims to the Holy Land
might recall that, for a thousand years, it was part of the Greek world, for
seven hundred years was ruled by Rome, and for three hundred years was
part of a Christian Greek world.
When we look beyond the Euphrates, the historical framework is profoundly different: classical writers offer representations of societies there
which were in various ways ‘‘foreign’’ and ‘‘oriental.’’ Alexander’s conquests,
which marked out the area of which the classical world knew and had continuing contact with, extended to regions characterised by different language groups, social structures, religious systems, and cultures. First was
Mesopotamia, particularly the southern part, Babylonia, heir to a long cultural tradition, persisting under successive empires. Writing in Akkadian,
using the cuneiform writing system developed from the late fourth millen. See M. Stern, Greek and Latin Authors.
. See the major work by S. Mitchell, Anatolia: Land, Men and Gods in Asia Minor I–
II ().
. F. Millar, The Roman Near East,  ..–..  ().

Rome and the East
nium, persisted throughout the Hellenistic period into the first century ..;
the latest cuneiform tablet so far known dates to .. –, and is thus contemporary with Vespasian.20 Second, further east, was the vast Iranian plateau, the culture of which is immensely difficult to characterise; it remains
an empty space on the map of the cultures of Asia. Third, beyond Iran, was
the region classical writers called ‘‘Arachosia’’ and ‘‘Bactria,’’ roughly modern
Afghanistan, Uzbekistan, and Tajikistan, in which scattered evidence reveals
a powerful imprint of Greek imperialism and colonialism. Finally, there was
India (primarily the Indus Valley, now in Pakistan), which attracted more
attention and stimulated more ethnographic—or pseudo-ethnographic—
writing in Greek and Latin than the other three regions combined: Herodotus and other Greek writers had spoken of India even before Alexander’s
expedition. The high profile of India was partly the legacy of the detailed accounts left by writers of Alexander’s time and of the early Hellenistic period;
and it was partly explained by the existence under the Roman Empire of
well-established trade routes from the Mediterranean to India, by sea from
Egypt, or by a combination of land and sea through Mesopotamia. The sight
of a ship setting sail for India had made Trajan think longingly of following
in the footsteps of Alexander.
A survey of the relations between the Greek and Roman world and this series
of regions should begin by recalling that Alexander had travelled south-east
across the Fertile Crescent to Babylonia, Susa, and Persis, then north and
north-east to Media and Bactria, and then south again to the Indus Valley.
With what conceptions, and with what long-term goals (if any) Alexander
and his forces had approached the cultures and societies they encountered,
it is hardly possible now to know. It may not be an accident, however, that
the clearest and most emphatic attribution of a motive to Alexander in classical literature comes from centuries later, and dates to precisely the period of
Trajan’s campaigns in the East. At that time, Plutarch wrote his On the Fortune of Alexander, in which he portrayed Alexander as bearing the benefits of
Greek civilisation to the barbarians:
But if you examine the results of Alexander’s instruction, you will
see that he educated the Hyrcanians to respect the marriage bond and
taught the Arachosians to till the soil, and persuaded the Sogdians to
support their parents, not to kill them, and the Persians to respect their
mothers and not to marry them. . . . When Alexander was civilising
. See J. Oelsner, Materialien zur babylonischen Gesellschaft und Kultur in hellenistischer Zeit
(), .
Looking East from the Classical World

Asia, Homer was commonly read, and the children of the Persians, of
the Susianians and of the Gedrosians learnt to chant the tragedies of
Sophocles and Euripides. . . . Thus Alexander’s new subjects would
not have been civilised, had they not been vanquished. Egypt would
not have its Alexandria, nor Mesopotamia its Seleucia, nor Sogdiana
its Prophthasia, nor India Boukephalia, nor the Caucasus (the Hindu
Kush) a Greek city near it.21
There is no doubt that Alexander, and later the Seleucids, did found many
Greek cities in Asia, though predominantly in Syria, Mesopotamia, and Bactria.22 Those that can be identified in the Iranian plateau are located along
the north-eastern edge, like Laodicea in Media; in the south, between the
Zagros Mountains and Mesopotamia proper, like Seleucia on the Eulaeus,
the ancient Susa; or (probably) on the Persian Gulf, like Antiochia in Persis. The plateau itself, in this respect as in so many others, remains largely a
blank. Plutarch’s idealising conception that a ‘‘civilising mission’’ had characterised Alexander’s campaign and that Greek culture and social norms had
been communicated to a variety of Asian societies should not be allowed
to colour our analysis of his actions.23 None the less, even Plutarch, though
imperialist in his belief in the superiority of Greek culture, does not speak
in racialist terms. Greek writing on Asia is not marked by particular attention to physical type or skin colour, or by the attribution of inherent racial
characteristics.
If racialism was not a significant factor, both imperialism and colonialism
were. Whatever the uncertainties in the evidence, we can be certain that a
number of ‘‘Alexandrias’’ and other Greek cities were founded in central Asia,
from Alexandria Eschate (‘‘the furthest’’) in ancient Sogdiana (Tajikistan) to
‘‘Alexandria in Arachosia’’ (Kandahar). One of these Greek cities, whose ancient name is unknown, and whose remains were discovered at a site called
Ai Khanum on the banks of the Amu Darya (the Oxus) in Afghanistan, was
extensively excavated from  until war brought work to a halt in .24
. Plutarch, Moralia C–F (trans. Loeb, vol. IV, –).
. For a very careful review of the complex and confusing evidence for city foundations
by Alexander himself in Asia, see P. M. Fraser, Cities of Alexander the Great (). See also
G. M. Cohen, The Seleucid Colonies: Studies in Founding, Administration and Organisation ().
. See especially P. Briant, ‘‘Colonisation hellénistique et populations indigènes: la
phase d’installation,’’ in his Rois, Tributs et Paysans (), ff.
. Perhaps the best sketch of what is known of the site, from which the discoveries
are still being studied, is P. Bernard, ‘‘Aï Khanum on the Oxus: A Hellenistic City in Central Asia,’’ Proc. Brit. Acad.  (), . For a more up-to-date survey, C. Rapin, ‘‘Greeks

Rome and the East
With almost too perfect an appropriateness, one of the Greek inscriptions
found there contains a set of maxims copied from those set up at Delphi, five
thousand kilometres to the west.25
The evidence hardly allows us to see what happened to these small, distant Greek cities in rapidly changing circumstances. We know that, in the
last years of the fourth century .., Seleucus I withdrew from northern
India and extensive areas to its north-west (‘‘Gedrosia’’ and ‘‘Arachosia’’ to
the Greeks) before the first king of the Mauryan dynasty, Chandragupta
(‘‘Sandrakottos’’ in Greek texts). In Bactria, further north-east, independent
or semi-independent Greek kingdoms arose in the third century .. and
continued until the second century. The scatter of small Greek inscriptions
from central Asia shows that the language remained in use (alongside Aramaic), without telling us whether it was used only by colonists, or by administrators using Greek, or whether its use had spread, as Plutarch’s words
suggest, widely among non-Greeks. Nothing prepares us for the official use
of Greek by Chandragupta’s grandson, the renowned king Asoka (probably – ..), a convert to Buddhism, whose great series of moralizing edicts, carved on rock faces or pillars over almost all of India, from
south to north, represents the earliest written evidence for the history of the
sub-continent.26 Even though we cannot know who precisely was being addressed, one is still amazed, forty years after the first of the Greek edicts was
published, to read Asoka’s message as expressed in two Greek inscriptions
discovered in Kandahar.
The first edict is accompanied by an Aramaic version, an important fact
in itself, and reads as follows:
Now that ten years have been completed, the king Piodasses [Asoka]
has demonstrated piety to men, and from that time on has made men
more pious, and everything prospers over all the land, and the king abstains from [eating] living things, and the other men, even such as are
the king’s hunters and fishermen, have ceased to hunt, and if any are
uncontrolled they have ceased from their excess as far as they can, and
in Afghanistan: Aï Khanum,’’ in J. Descoeudres, ed., Greek Colonists and Native Populations
(), .
. See the evocative paper by L. Robert, ‘‘De Delphes à l’Oxus: inscriptions grecques
nouvelles de la Bactriane,’’ CRAI (): .
. See R. Thapar, Aśoka and the Decline of the Mauryas 2 (), with two maps at the end
showing the distribution of the inscribed edicts. See F. Millar, ‘‘Epigraphy,’’ in M. Crawford, ed., Sources for Ancient History (), – ( chapter  in F. Millar, Rome, the Greek
World, and the East I: The Roman Republic and the Augustan Revolution, on p. ).
Looking East from the Classical World

[are] obedient to their fathers and mothers and their elders to a degree not seen before; and for the future, acting in accordance with all
these principles they will conduct themselves in a more desirable and
better way.27
Less attention has been paid to the second inscription of the pair, published
in ,28 a remarkable text only once reproduced for study by classical historians, and none of it, in its Greek version, ever translated into English.29
The text, containing the end of Edict XII of Asoka and the beginning of
Edict XIII—in which he expresses his remorse over the slaughter in the recent war in Kalinga—appears, in this case, only in Greek and has been inscribed on a building block. As Daniel Schlumberger observes, it is an adaptation rather than a literal translation of the Indian versions which are known
from elsewhere.
It may be sufficient to translate here the first few lines of Edict XIII, referring to the Kalinga war:
While Piodasses was in the eighth year of his reign, he conquered
Kalinga. There were captured and deported from there , persons, and another , were killed, and almost as many others died.
From that moment remorse and pity took hold of him, and he was
grieved. In the same way as he ordered that men should abstain from
[eating] living things, he exercises zeal and effort in the pursuit of piety
. First published by D. Schlumberger, L. Robert, A. Dupont-Sommer, and E. Benveniste, ‘‘Une bilingue gréco-araméenne d’Asoka,’’ Journal Asiatique  (): ; text reproduced in Supplementum Epigraphicum Graecum XX (), no. , and in J. Pouilloux,
Choix d’inscriptions grecques (), no. , with French translation. An English translation
can be found also in S. M. Burstein, The Hellenistic Age from the Battle of Ipsos to the Death of
Kleopatra VII (), no. .
. E. Benveniste, ‘‘Edicts d’Asoka en traduction grecque,’’ Journal Asiatique  ():
; D. Schlumberger, ‘‘Une nouvelle inscription grecque d’Açoka,’’ CRAI (): . I
have not seen D. Schlumberger and E. Benveniste, ‘‘A New Greek Inscription of Asoka at
Kandahar,’’ Epigraphia Indica  (): . French translation in J.-M. Bertrand, L’Hellénisme
– av. J.-C. (), –.
. For what seems to be the sole reprinting of the Greek text, see R. Schmitt, ‘‘EX
OCCIDENTE LUX. Griechen und griechische Sprache im hellenistischen Fernen Osten,’’
in P. Steinmetz, ed., Beiträge zur hellenistischen Literatur und ihrer Rezeption in Rom (),
, on –, with German translation. Although an English translation of the relevant
edicts (XII and XIII), derived from various versions in Prakrit, is provided in Thapar, Aśoka
(n. ), –, this part of the book is a reprint of the original  text, and the translation
therefore does not depend on the Greek version.

Rome and the East
[eusebeia]. And this the king has felt with still greater displeasure; [that?]
such as dwelt there who were Bramenai or Sramenai, or also any others
who spend their lives on piety . . . if any of these following such a way
of life has died or been deported . . .
To a study of the classical world’s relationship with Asia, the edicts and minor
inscriptions of Asoka are of greater significance than might be supposed.
It is striking enough to read an expression of Buddhist doctrines in Greek.
With the exception of the undeciphered Indus Valley Script of the third and
early second millennium, the inscriptions of Asoka are, in fact, the earliest
written documents from the Indian sub-continent. Apart from Greek and
Aramaic, the edicts were written in a set of dialects related to Sanskrit, and
called Prakrit; the balance of scholarly opinion is that the two scripts used
for the Prakrit versions, namely Kharosthi (confined to the north-west, and
always written right to left) and Brahmi (normally written left to right), both
derived from Aramaic.30 Thus, both the languages and scripts used by Asoka
suggest that monumental writing in India owed its origins to the influence
of the Achaemenid Empire, in which Aramaic was widely used as an official language, and of the Seleucid Empire, in which Aramaic was still used
alongside the primary language, Greek.
Elsewhere in his edicts (in the so-called Fifth Major Rock Edict), Asoka
does indeed list Greeks as being among the peoples of his empire, and he also
refers to ‘‘the Greek king named Antiochus’’ (Antiochus II, – ..). It
is unfortunate that the Greek text of Edict XIII breaks off where it does, for
here, according to the Prakrit versions, Asoka goes on to refer again to Antiochus, and also to other Hellenistic kings: ‘‘and beyond the realm of that
Antiochus, in the lands of the four kings named Ptolemy, Antigonus, Magas
and Alexander . . .’’ 31
No hint of interaction between Greek thought and Buddhism appears in
any pagan classical source, and, in the surviving Greek literature, the earliest
mention of the Buddha by name is by the Christian writer Clement of Alexandria at the end of the second century.32 Classical narratives of the history of
. See R. G. Salomon, ‘‘Brahmi and Kharosthi,’’ in P. T. Daniels and W. Bright, eds.,
The World’s Writing Systems (), .
. Thapar, Aśoka (n. ), . The kings are Ptolemy II Philadelphus (– ..),
Antigonus Gonatas (– ..), Magas of Cyrene (dates not quite certain), and (probably) Alexander of Epirus (– ..).
. Clement, Stromata , , : ‘‘There are among the Indians those who believe in the
teachings of Boutta, whom because of his exceptional nobility they have honoured as a
Looking East from the Classical World

the Greek successor states which replaced the Seleucids in the Bactrian area,
and lasted into the second century .., are also very slight. It is not worth
summarizing here the complex and uncertain sequences of rulers and dynasties which can be worked out, primarily from their exceptionally beautiful
and impressive coins with Greek legends.33 Polybius’ narrative of Hellenistic
history, fragmentary as it is, does touch on this area, however, in connection
with the march into Central Asia by the Seleucid king Antiochus III, ‘‘the
Great,’’ in the last years of the third century .. Polybius represents one of
the local Greek kings, Euthydemus, as defending his usurpation of power on
the grounds that Seleucid control of the region had broken down earlier,
and that threats by nomad invaders meant that if the area were not defended
it would soon be ‘‘barbarised’’ (ekbarbarōthēsesthai).34
Greek rule in Bactria did in the end disappear, though not as immediately as Euthydemus feared. One of the most striking revelations of the past
decade is the publication of the first perishable document discovered in a
Greek kingdom in Bactria, a Greek parchment, apparently of the first half
of the second century .. The text, fragmentary as it is, reflects a fully Hellenised regime: ‘‘In the joint reign of the God Antimachus and of Eumenes
and Antimachus . . . year , month Olous, in Asangorna when [. . .] was
guardian of the law [nomophylax], Menodotus the tax-gatherer [logeutēs] . . .’’ 35
Later in the century, Bactria was conquered by the Sakas or Shakas, whose
kings continued to use Greek titles and Greek coin legends. In the meantime, however, Greek rulers, who naturally continued to use Greek on their
coinage, had extended their control south of the Hindu Kush mountains to
god.’’ See A. Dihle, ‘‘Indische Philosophen bei Clemens Alexandrinus,’’ in his Antike und
Orient: Gesammelte Aufsätze (), , and G. M. Bongard-Levin and S. Karpyuk, ‘‘Nachrichten über den Buddhismus in der antiken und frühchristlichen Literatur,’’ in B. Funck,
ed., Hellenismus: Beiträge zur Erforschung von Akkulturation und politischer Ordnung in den Staaten
des hellenistischen Zeitalters (), .
. See esp. the classic work of W. W. Tarn, The Greeks in Bactria and India 3 (), with
preface and bibliography by F. L. Holt; see also R. K. Narain, The Indo-Greeks (). For a
fine selection of these royal coins, see N. Davis and C. M. Kraay, The Hellenistic Kingdoms:
Portrait Coins and History (), chap. .
. Polybius ,  (Loeb ed., vol. IV, –).
. Published independently by J. R. Rea, A. S. Hollis, and R. C. Senior, ‘‘A Tax Receipt from Hellenistic Bactria,’’ Zeitschr. f. Pap. u. Epig.  (): , and by P. Bernard
and C. Rapin, ‘‘Un parchemin gréco-bactrien d’une collection privée,’’ CRAI (): .
See C. Rapin, ‘‘Nouvelles observations sur le parchemin gréco-bactrien d’Asangôrna,’’ Topoi
 (): .

Rome and the East
the northern part of the Indian sub-continent. It was only in the first century .. that both this area and Bactria fell under the rule of the Kushans,
who also initially used Greek for their coin legends.36 John Boardman’s study
of the diffusion of Greek art shows how profound its influence still was, for
instance, on ‘‘Gandharan’’ sculpture and on the earliest sculptural representations of the Buddha, which date to the first century ..37 Perhaps the most
striking item is the coin of the Kushan king Kanishka, probably dating to
the early second century .., which bears a representation of the Buddha
on the reverse, accompanied by the Greek legend ΒΟΔΔΟ.38
The reign of Kanishka offers some of the most remarkable of all the evidence for the profound and long-lasting influence of Greek in Afghanistan
and northern India. His coins show a transition from the use of royal titles
in Greek to ones in an Iranian language conventionally labelled ‘‘Bactrian.’’
But the script used continues to be Greek, with the addition of one letter,
shaped somewhat like a Greek rho (Ρ ), but with the vertical stroke continued
upwards, which was used for the sound ‘‘sh.’’ A truly remarkable find is an
extensive inscription of the first year of Kanishka’s reign, found in  at
Rabatak in Afghanistan, and written in Bactrian using this version of the
Greek alphabet. The text, proclaiming the king’s assumption of power, with
divine favour, and recounting the cities in northern India which he ruled,
records explicitly that ‘‘Kanishka the Kushan’’ (Kanēshke koshano) ‘‘issued a
Greek edict and [then] put it into Aryan.’’ A substantial group of Bactrian
documents, one written on copper and the others on perishable materials,
dating to the fourth to eighth century, and still using a cursive variant of the
same (more or less) Greek alphabet, awaits publication.39 The edict is evidence of the length of both Greek and Iranian influence in the region. Perhaps only the great inscription of the second Sassanid king, Shapur I (who
. For an illuminating survey, with extensive illustrations, see E. Errington and J. Cribb,
eds., The Crossroads of Asia: Transformation of Image and Symbol in the Art of Ancient Afghanistan
and Pakistan ().
. J. Boardman, The Diffusion of Classical Art in Antiquity (), chap. , ‘‘The East after
Alexander.’’
. Boardman (n. ),  (coin reproduced on ); Errington and Cribb (n. ), –
.
. N. Sims-Williams and J. Cribb, ‘‘A New Bactrian Inscription of Kanishka the Great,’’
Journal of the Institute of Silk Road Studies  (–): ; N. Sims-Williams, ‘‘Nouveaux documents sur l’histoire et la langue de la Bactriane,’’ CRAI ():  (also reporting on the
later documents). For the first group of these documents, see N. Sims-Williams, Bactrian
Documents from Northern Afghanistan, Pt. : Legal and Economic Documents ().
Looking East from the Classical World

reigned in .. –), surpasses it as testimony to the long-lasting importance of Alexander.
As we have reached the period when both a well-established sea route and a
land-sea route linked the Mediterranean with India, it is time to turn back
to examine the (paradoxically) less well attested, and more mysterious, relations of the Graeco-Roman world with Iran. The area is geographically closer
to the Mediterranean, and its successive rulers either were (for a time) themselves Greek, during the first part of the Hellenistic period, or alternatively
were in close contact, and often conflict, with the Graeco-Roman powers, as
was the case with Achaemenid Persia, Parthia, and Sassanid Persia. The mystery arises out of the lack of contemporary evidence in any Iranian language.
This is true of the Achaemenid Persian Empire, from which the only strictly
Persian documents are royal inscriptions in Old Persian, written in cuneiform, as were Achaemenid documents in Elamite or Akkadian. Whether
there was also an alphabetic (or any other) system for writing Old Persian on
perishable materials is a matter of speculation; the thousands of administrative documents known from Persepolis are written mainly in Elamite, but
also in Aramaic.
There is an equal silence for the period of Hellenistic control of Iran, and
an almost equally complete one from the Parthian period. It was of great political and strategic significance that a little-known Iranian people, the Parthians, replaced the Seleucids in Iran in the early second century .. and
extended their empire to Babylonia and the middle Euphrates later in the
century. The Parthian Empire was to be the main adversary of the Romans
in the Near East; Trajan had marched against them on the campaign which
took him to the Persian Gulf. Although there is some documentation written
on perishable materials or on fragments of pottery, and some coin legends,
to show that Parthian, an Iranian language, could be written using the Aramaic alphabet,40 not a single item of literature in an Iranian language survives
from the Parthian period, which lasted for more than four centuries until the
s .., when the Parthian kings were replaced by a new Persian dynasty
originating from Fars (Persis) in the south of Iran, the Sassanids.
The Parthians ruled over much the same area as the Seleucids, except that
they never conquered Syria, and within their territories Aramaic was widely
. See P. V. Skjaervo, ‘‘Aramaic Scripts for Iranian Languages,’’ in Daniels and Bright
(n. ), . The largest group of Parthian texts is represented by more than , ostraca
from Nisa in Turkmenistan; see I. Diakonoff and V. Livshits, Corpus Inscriptionum Iranicarum II.: Parthian Economic Documents from Nisa I–III, ed. D. Mackenzie (–).

Rome and the East
used, as it had been under the Achaemenids and the Seleucids. Akkadian survived in Babylonia at least until towards the end of the first century ..,
and Greek cities, deriving from early Hellenistic colonisation, like Seleucia on the Tigris and Seleucia on the Eulaeus—the ancient Persian city of
Susa in Elymais (Elam)—flourished in the Parthian Empire. There was also
the (now) famous small Hellenistic settlement of Dura-Europos on the Euphrates, which was excavated, if not exactly by, at any rate under the direction of, Franz Cumont and Mikhail Rostovtzeff in the s and s.
Thus, the Parthian Empire was a multi-cultural, or multi-ethnic, world in
which the culture least visible to us was that of the Parthians themselves. Although the original documentation surviving from the region in this period
is scanty, a significant amount of what there is was written in Greek. In other
words, the Parthian Empire was, in one respect, just another late Hellenistic
kingdom.
Take, for instance, the two long Greek parchments of the first century ..
(– and – .., respectively), whose origin E. H. Minns, when publishing them in , describes in terms which now seem striking: ‘‘Avroman
is a town in Persian Kurdistan lying close to the Turkish Frontier.’’ 41 In fact,
the place of origin lies in the foothills of the Zagros, north-north-east of
Baghdad, in present-day Iraq. The dating of the earlier of the two parchments perfectly reflects the character both of the Parthian kings as late Hellenistic monarchs and also the power relations of the time: ‘‘In the reign of
the King of Kings, Arsaces, benefactor, just, manifest, philhellene, and of
the Queens, Siace his compaternal sister and wife and Aryazate surnamed
Automa, daughter of the great King Tigranes (of Armenia) and his wife . . .’’
Next, one might cite the inscription of ..  containing a letter in Greek
from the Parthian king, Artabanus III, addressed to the city of Susa on the
Eulaeus;42 or, perhaps more indicative, a series of nine business and legal
documents written in Greek on parchment, found in Dura-Europos and
dating from the last eighty years of Parthian rule, before the Roman advance
of the s.43 Although Dura-Europos was a Greek settlement, this evidence
. E. H. Minns, ‘‘Parchments of the Parthian Period from Avroman in Kurdistan,’’ Journ.
Hell. Stud.  (): . It seems extraordinary that these two extensive Greek documents,
along with one in Parthian, now in the British Museum, have never been re-studied in
detail since.
. C. B. Welles, Royal Correspondence in the Hellenistic Period (), no. .
. H. M. Cotton, W. E. H. Cockle, and F. G. B. Millar, ‘‘The Papyrology of the Roman
Near East: A Survey,’’ Journ. Rom. Stud.  (): , nos.  and  to . See also F. Millar, ‘‘Dura-Europos under Parthian Rule,’’ in J. Wiesehöfer, ed., Das Partherreich und seine
Zeugnisse (Historia-Einzelschrift , ), – ( chapter  in the present volume).
Looking East from the Classical World

for the conduct of public legal business there in Greek none the less reveals
the workings of the Parthian Empire. Equally striking in a different way is
the bilingual text from the mid-second century .. inscribed on a statue of
Herakles to record that the Parthian king Vologaeses IV had taken it as part
of the spoils when he captured Mesene at the head of the Persian Gulf and
brought it back to Seleucia on the Tigris. The Greek text is accompanied by
one in Parthian, written as always in Aramaic letters.44
Although Greek was a normal and established language for public documents in the Parthian Empire, the Iranian languages—first Parthian, then
Persian proper, then others (apart from ‘‘Bactrian,’’ which used Greek letters)—borrowed from Aramaic, the alphabet which they were to use for
written texts. This complex evolution is dramatically illustrated by the earliest document of the Sassanid Persian Empire, which replaced the Parthian
Empire in the s. Shapur I left at Naqsh-i Rustam near Persepolis in southern Iran a long public record of his military and civil achievements, inscribed
in Greek, Parthian, and Middle Persian, the latter two again using the Aramaic alphabet, written from right to left.45 The vast Greek text in which
Shapur describes his wars against Rome in the s and s and explains the
principles of his civil government may be regarded as the ultimate testimony
to the long-lasting legacy of Alexander’s conquests six hundred years earlier.
The late borrowing of an alphabetic script for writing Persian is of crucial
importance, as is the fact that for the whole of the Sassanid period, which
lasted until the Islamic conquests of the seventh century, we have no contemporary manuscripts of works written in Persian or other Iranian languages, except for Manichaean manuscripts from central Asia.46 There are
literary works of the Islamic period which are written in Persian, and which
are claimed to have been compiled in the Sassanid period, or to be based on
works of that period, but there are no actual manuscripts. We might contrast
. Greek text in Supplementum Epigraphicum Graecum XXXVII, no. . See esp. P. Bernard, ‘‘Vicissitudes au gré de l’histoire d’une statue en bronze d’Héraclès entre Séleucie du
Tigre et la Mésène,’’ Journal des Savants (): .
. See A. Maricq, ‘‘Res Gestae Divi Saporis,’’ Syria  (): , and now Ph. Huyse,
Die dreisprachige Inschrift Sabuhrs I an der Kala-i Zardust (). See the invaluable examples
of the Parthian and Middle Persian scripts used in this inscription which are provided by
Skjaervo (n. ), –. An English translation is given in R. N. Frye, The History of Ancient
Iran (), –.
. See the careful study of the evidence by J. Wiesehöfer, Das antiken Persien von 
v. Chr. bis  n. Chr. (), – (a survey of the epigraphic and literary evidence for
the Sassanid period); English translation by A. Z. Azodi, Ancient Persia from  to  ..
(), –.

Rome and the East
with this silence the long list of ancient papyri with texts, fragmentary or
extensive, of Greek (and some Latin) literary works, or the striking series of
dated Syriac manuscripts, beginning with one written in Edessa in .. ,
now in the British Museum.47 From Iran there is nothing comparable.
Out of this silence emerges the most remarkable case of the construction,
or representation, of an ‘‘oriental’’ religious system by Greek, and later also
Roman, writers: Zoroastrianism. It is not claimed here that we can know that
the conventional account of Zoroaster, of his teaching, and of its diffusion
in Iran is a fiction; only that, so far as evidence from antiquity is concerned,
our (supposed) knowledge of Zoroastrianism depends entirely, and without exception, on Graeco-Roman representations. However, this view is not
conceded in the standard works of reference. P. V. Skjaervo, speaking of the
central text of medieval and modern Zoroastrianism, the Avesta, remarks:
‘‘The oldest Avestan texts are approximately contemporary with the Rigveda
(second millennium ...), while the younger texts date from the first millennium ...; they were transmitted orally and written down only in the
mid-Sasanian period, the fifth–sixth centuries .. (though the oldest manuscripts date only from the th century).’’ 48
The history of Zoroaster is, in fact, derived solely from claims made within
the Zoroastrian texts themselves; the earliest references to Zoroaster contained in texts in Iranian languages are found in Manichaean writings discovered in central Asia. It is exceedingly difficult to date these manuscripts,
or the works which they contain, but none can be earlier than the lifetime of
Mani himself, the third century .., and most must be much later.49 The specifically Zoroastrian texts comprise a set of religious writings of the Islamic
period which project the figure of Zoroaster, and the story of the Avesta as
a text, back into the remote past.
The ancient evidence for Zoroaster, his date, and his teachings is all, without exception, Graeco-Roman. It begins with Xanthos the Lydian in the late
fifth century .., who claimed that Zoroaster lived six thousand years before Xerxes invaded Greece, and continues throughout the classical period.50
. See the splendid collection by W. H. P. Hatch, Album of Dated Syriac Manuscripts
(), which could now be substantially updated.
. Skjaervo (n. ), .
. For these texts in English translation, see H. J. Klimkeit, Gnosis on the Silk Road:
Gnostic Texts from Central Asia (). Note, e.g., the ‘‘Zarathustra Fragment’’ (p. ), and the
allusion to Zarathustra in a poem in Parthian on p. . No date is suggested for either text
or either manuscript.
. The texts are conveniently collected and discussed by J. Bidez and F. Cumont, Les
Looking East from the Classical World

Classical sources provide by far the earliest representations of Zoroaster as an
Iranian sage located in the remote past, whose teachings expressed a dualistic view of the world. The most informative of these texts is a passage from
Plutarch’s essay On Isis and Osiris, a work written in the early second century ..:
The great majority and the wisest of men hold this opinion: they believe that there are two gods, rivals as it were, the one the Artificer of
good and the other of evil. There are also those who call the better one
a god and the other a daemon, as, for example, Zoroaster the sage, who,
they record, lived five thousand years before the time of the Trojan
War. He called the one Oromazes and the other Areimanius; and he
further declared that among all things perceptible to the senses Oromazes may best be compared to light, and Areimanius, conversely, to
darkness and ignorance, and midway between the two is Mithras; for
this reason the Persian give to Mithras the name of ‘‘Mediator.’’ 51
The conclusion is obvious. No problem arises from Plutarch’s reference to
the major Iranian deity, Ahura Mazda, for the great inscription of Darius I,
from the late sixth century .., written in old Persian, proclaims the king
as his worshipper. Neither in this long inscription, however, nor anywhere
else in the scanty Iranian evidence from before the early medieval period, is
there any allusion to an ancient teacher named Zoroaster.
The reader might reasonably object that the manuscripts of Plutarch’s
works, too, are medieval. So they are. The fundamental difference, however,
is that the texts of classical authors such as Plutarch, while mainly derived
from medieval manuscripts, can be tied into a dense interlocking network
of ancient papyrus texts and of ancient documents, both papyri and inscriptions, some naming Plutarch himself, as well as many other persons who
appear in his Lives and Moral Essays. In this case, therefore, we can work back
from the medieval copy to a known ancient context.With the Avesta, we cannot. A series of classical writers ascribe to a very ancient Persian sage named
‘‘Zoroaster’’ a dualistic view of the world. We cannot prove them wrong, but
it would be counter to all methodological principles to treat such representamages hellénisés I–II (). See now A. De Jong, Traditions of the Magi: Zoroastrianism in Greek
and Latin Literature (). Note also P. Kingsley, ‘‘The Greek Origin of the Sixth-Century
Dating of Zoroaster,’’ Bull. Sch. Or. and Afr. Stud.  (): , vigorously criticising the
Greek tradition—but only (it seems) in favour of a second-millennium date for the ‘‘real’’
Zoroaster.
. Plutarch, Moralia D–D (trans. Loeb text, vol. V, –).

Rome and the East
tions as fac